HEDGEFUNDIE's excellent adventure Part II: The next journey

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
quicklink7
Posts: 2
Joined: Fri Nov 27, 2020 7:29 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by quicklink7 »

Could anyone please explain to me how interest rate changes effect the leveraged stock/s (swap?) since this is borrowing money to invest? i believe it was the libor rate but what happens if interest rates go to 10% or 15%? Will the yearly % gain be reduced by only 15% if interest rates went that high? Since i believe it uses swaps at a libor rate if the rate went higher would the stocks gains also take a hit and if so how much? Or would it not impact the stock that much?

Thank you.
hell0men
Posts: 4
Joined: Sun Nov 01, 2020 12:53 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by hell0men »

Raraculus wrote: Fri Nov 27, 2020 3:41 pm

I'm thinking of just investing in SSO alone and use whatever cashflow I have to paper over the dips. Backtesting (past 12 months) this strategy did reasonably well. Sure, it was volatile, but nothing like UPRO. I could actually keep at this strategy for a whole year, just using my regular contributions.

When the market goes down, the fund with the leverage of x3 goes down more and we buy it instead of x2 or x1, increasing the leverage. UPRO / SSO / SPY or TQQQ / QLD / QQQ .
In the initial set of positions, you can keep the ETF without the leverage and increase the leverage gradually by smearing the risks, because by buying UPRO at once we risk right from the start to fail in a long drawdown.
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

djeayzonne wrote: Fri Nov 27, 2020 7:00 pm
Hello, I wanted to revisit this discussion.
Why 10 deltas and 120 days out resulting in very cheap options requiring you to buy so many?
I would think the commissions on doing this would make this SIGNIFICANTLY more expensive to actually implement.

For that reason, why not 60 days out and a strike price of 40% over current VIX levels, for example?
Or 30 deltas instead of 10?
5,000 VIX options costs $89 to enter the position where I trade. Zero to exit if you don't cash them. If you cash them it would be $39 in clearing and regulatory fees.

The VIX hedge is preemptive. You don't cash it often but when you do you then you want the maximum number of options to benefit from the options moving from extrinsic to intrinsic value (hence the preference for the lower delta. We are benefitting from tail risk as buyers so the higher number of options owned that move to ITM the better the payout.
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

cooljack4u wrote: Fri Nov 27, 2020 6:43 pm Does it make sense to have rules as when to hedge and when to cash out ? I understand the insurance concept. Is there a way to reduce the cost have the same benefit. What if we but the VIX options only in the months when VIX is below 15 ? Also, is there a optimal cashing out number ? like you sell the options when VIX is crossing 40 and invest that money back into TQQQ ?
I want to cash my VIX calls when they are over $60 as a minimum threshold. Prefer to get strikes below $50 when buying the calls. If I can't get them at 10 delta then I'll rather buy shorter days to expiration contracts as long as backwardation is not in effect.

Sure it's great to get the calls when it's below 15, but the VIX is currently at 20.84. What if the market craters in the next couple of weeks? No one knows when the black swans are arriving. 2 weeks or 10 years?
djeayzonne
Posts: 104
Joined: Wed Dec 06, 2017 2:14 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by djeayzonne »

perfectuncertainty wrote: Sat Nov 28, 2020 5:12 am
djeayzonne wrote: Fri Nov 27, 2020 7:00 pm
Hello, I wanted to revisit this discussion.
Why 10 deltas and 120 days out resulting in very cheap options requiring you to buy so many?
I would think the commissions on doing this would make this SIGNIFICANTLY more expensive to actually implement.

For that reason, why not 60 days out and a strike price of 40% over current VIX levels, for example?
Or 30 deltas instead of 10?
5,000 VIX options costs $89 to enter the position where I trade. Zero to exit if you don't cash them. If you cash them it would be $39 in clearing and regulatory fees.

The VIX hedge is preemptive. You don't cash it often but when you do you then you want the maximum number of options to benefit from the options moving from extrinsic to intrinsic value (hence the preference for the lower delta. We are benefitting from tail risk as buyers so the higher number of options owned that move to ITM the better the payout.
Where is that?
The only broker even close would be tasty works with their 10 dollar cap per leg, but even that is up to 250 contracts.
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

djeayzonne wrote: Sat Nov 28, 2020 5:37 am
Where is that?
The only broker even close would be tasty works with their 10 dollar cap per leg, but even that is up to 250 contracts.
I have an account with Tasty Works. The numbers I quoted you are for $50 contracts (5,000 options). Tasty Works does not cap commissions on index options.

BTW - Tasty Works equity options there is no 250 contract cap. It's $10 per leg. The 250 contract order limit can be upped with a phone call - it’s the default limit in the platform.

Thinkorswim is actually cheaper on index options contracts since they clear their own trades versus Tasyworks who use Apex to clear for them. The retail commission and fees for 5,000 options (50 cars) is $32.50 commission + $9.20 in fees ($41.70). If they expire worthless that’s the total cost. If you cash them it's another $41.70. In total cheaper than Tastyworks. If you trade options with enough frequency and volume TOS will negotiate on commissions below the figures I quoted above.

The prices I quoted are actual quotes from each of the platforms.
Raraculus
Posts: 60
Joined: Sat Jul 20, 2019 10:43 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Raraculus »

hell0men wrote: Sat Nov 28, 2020 12:49 am
Raraculus wrote: Fri Nov 27, 2020 3:41 pmI'm thinking of just investing in SSO alone and use whatever cashflow I have to paper over the dips. Backtesting (past 12 months) this strategy did reasonably well. Sure, it was volatile, but nothing like UPRO. I could actually keep at this strategy for a whole year, just using my regular contributions.
When the market goes down, the fund with the leverage of x3 goes down more and we buy it instead of x2 or x1, increasing the leverage. UPRO / SSO / SPY or TQQQ / QLD / QQQ .

In the initial set of positions, you can keep the ETF without the leverage and increase the leverage gradually by smearing the risks, because by buying UPRO at once we risk right from the start to fail in a long drawdown.
That's pretty insightful comment you've got there! Made me rethink my position regarding leveraged ETF's. Thank you.
djeayzonne
Posts: 104
Joined: Wed Dec 06, 2017 2:14 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by djeayzonne »

perfectuncertainty wrote: Sat Nov 28, 2020 11:32 am
djeayzonne wrote: Sat Nov 28, 2020 5:37 am
Where is that?
The only broker even close would be tasty works with their 10 dollar cap per leg, but even that is up to 250 contracts.
I have an account with Tasty Works. The numbers I quoted you are for $50 contracts (5,000 options). Tasty Works does not cap commissions on index options.

BTW - Tasty Works equity options there is no 250 contract cap. It's $10 per leg. The 250 contract order limit can be upped with a phone call - it’s the default limit in the platform.

Thinkorswim is actually cheaper on index options contracts since they clear their own trades versus Tasyworks who use Apex to clear for them. The retail commission and fees for 5,000 options (50 cars) is $32.50 commission + $9.20 in fees ($41.70). If they expire worthless that’s the total cost. If you cash them it's another $41.70. In total cheaper than Tastyworks. If you trade options with enough frequency and volume TOS will negotiate on commissions below the figures I quoted above.

The prices I quoted are actual quotes from each of the platforms.
Phew!!! Glad you cleared that up! And, thanks for clarifying how tasty works works.

Yeah, actually I'm with IB and thinking to move to either TDA or Tastyworks. Negotiating with TDA at the moment.
djeayzonne
Posts: 104
Joined: Wed Dec 06, 2017 2:14 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by djeayzonne »

djeayzonne wrote: Sat Nov 28, 2020 5:30 pm
perfectuncertainty wrote: Sat Nov 28, 2020 11:32 am
djeayzonne wrote: Sat Nov 28, 2020 5:37 am
Where is that?
The only broker even close would be tasty works with their 10 dollar cap per leg, but even that is up to 250 contracts.
I have an account with Tasty Works. The numbers I quoted you are for $50 contracts (5,000 options). Tasty Works does not cap commissions on index options.

BTW - Tasty Works equity options there is no 250 contract cap. It's $10 per leg. The 250 contract order limit can be upped with a phone call - it’s the default limit in the platform.

Thinkorswim is actually cheaper on index options contracts since they clear their own trades versus Tasyworks who use Apex to clear for them. The retail commission and fees for 5,000 options (50 cars) is $32.50 commission + $9.20 in fees ($41.70). If they expire worthless that’s the total cost. If you cash them it's another $41.70. In total cheaper than Tastyworks. If you trade options with enough frequency and volume TOS will negotiate on commissions below the figures I quoted above.

The prices I quoted are actual quotes from each of the platforms.
Phew!!! Glad you cleared that up! And, thanks for clarifying how tasty works works.

Yeah, actually I'm with IB and thinking to move to either TDA or Tastyworks. Negotiating with TDA at the moment.
Having cleared the commission issue, I am still interested in why you chose the 10 Delta 120 days out. Please elaborate.
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

djeayzonne wrote: Sun Nov 29, 2020 3:39 am
Having cleared the commission issue, I am still interested in why you chose the 10 Delta 120 days out. Please elaborate.
Please refer to this spreadsheet as a reference LINK. Note that I added a tab with the VIX chart with some minor annotation.

I prefer to be mechanical in my approach when I execute, but I want a well thought out approach.

So here is a criterion and desired state description:

1. We don't know when a black swan is coming or how long it lasts.
2. We want a preemptive hedge in place for when it does arrive.
3. We want to benefit from tail risk events not be a victim to them.

Since I don't know when a black swan is coming and how long it might last I want to have the hedge in place at all times and I want it to be able to sustain the event. I want to be prepared.

Take a look at the 4-month hedge in the spreadsheet and the VIX Chart I referenced above.

Look at the chart first. The 4 magenta arrows on the left represent the 4 months where each leg of the hedge was purchased. This is described in the summary halfway down under VIX Hedge Preparation (November we pe purchase the 120 DTE 10 delta calls etc.).

Why 120 DTE?

Let's answer the "why 120 DTE" question first. Refer back to the chart and note the green arrows that show how long the spike in the VIX spanned. It started on Feb 21 and lasted till May 8th. In this case almost 3 months. What we know about VIX spikes is that they occur inversely to SP500 cliff dives. We do not know when they will peak and exactly how long they will last but they don't last very long. And once they start occurring it's almost too late to decide to buy the calls as they rocket so quickly. But we know we want to be ready BEFORE the VIX spike, and we want to have enough to span the event because we do not know when the VIX will peak.

To accomplish that we need multiple months of calls pre-purchased as cheaply as possible. Hence why I purchase them 120 DTE so that my options can span 4 months. One could possibly argue that we can do 90 DTE. As a point of interest, look at the chart and you will note that the peak occurred on March 18 which was exactly the date of expiration for March 2020. However, the highest closing price occurred on March 16. This is coincidental because what if it occurred a week later? Then you would want to be sure we had April calls in play too. The point for me is that we need multiple months to be in play. In 2008, the VIX started spiking in August and the spike lasted till December.

Why 10 deltas?

The reason why I buy 10 deltas comes down to the fact that when the VIX hedge spike it spikes big. I want to own as many call options as possible.. When the VIX spikes typically backwardation is occurring (front months are worth more than back months) and options that were 100% extrinsic value are quickly converted to intrinsic value.

In 2008, the VIX spiked to $96.40 on October 23 but closed at $67.80. That also was not the highest closing day. It closed over $80 the next day and over $79 the day after that. It retreated and spiked again to close over 80 again on November 20th.

Sorry for the ramble but it shows 2 things: 1. The VIX spike can last multiple months and we want to benefit from it as much as possible without being subject to a monthly expiration that occurs in the beginning and miss out on all the benefits of the spike. 2. True VIX spikes will easily envelope the lower deltas and I'll benefit from the greater number of calls because once an option becomes intrinsic the size of the move is constrained as the delta approaches 100.

As an example, I just pulled up the current VIX month and the table below makes the point for $1000 invested in 30 deltas versus 10 deltas:

Code: Select all

 Delta 	 Strike	Price 	Number of Options 
 30.00 	 28.00 	 1.10		  909.09 
 11.00 	 40.00 	 0.40		2,500.00 
So a difference of $12 in strike price but you get 2.75 times more options. In this example, if the VIX closed at $50 on expiration day the 30 deltas would be worth $18,810 whereas the 10 deltas would be worth $24,000 (27.6% more).

Hopefully, this helps answer why 120 DTE and why 10 deltas. As I stated earlier, I want to be mechanical in the execution wherever possible. Currently, it's not that easy since the VIX has remained elevated above historic levels this year so I've made some temporary adjustments.
User avatar
Ramjet
Posts: 663
Joined: Thu Feb 06, 2020 11:45 am
Location: Cleveland

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Ramjet »

For anyone who is using the target volatility approach with a look back, do you plan to move from TMF to something like EDV or cash if interest rates are rising consistently?
djeayzonne
Posts: 104
Joined: Wed Dec 06, 2017 2:14 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by djeayzonne »

perfectuncertainty wrote: Sun Nov 29, 2020 10:13 am
djeayzonne wrote: Sun Nov 29, 2020 3:39 am
Having cleared the commission issue, I am still interested in why you chose the 10 Delta 120 days out. Please elaborate.
Please refer to this spreadsheet as a reference LINK. Note that I added a tab with the VIX chart with some minor annotation.

I prefer to be mechanical in my approach when I execute, but I want a well thought out approach.

So here is a criterion and desired state description:

1. We don't know when a black swan is coming or how long it lasts.
2. We want a preemptive hedge in place for when it does arrive.
3. We want to benefit from tail risk events not be a victim to them.

Since I don't know when a black swan is coming and how long it might last I want to have the hedge in place at all times and I want it to be able to sustain the event. I want to be prepared.

Take a look at the 4-month hedge in the spreadsheet and the VIX Chart I referenced above.

Look at the chart first. The 4 magenta arrows on the left represent the 4 months where each leg of the hedge was purchased. This is described in the summary halfway down under VIX Hedge Preparation (November we pe purchase the 120 DTE 10 delta calls etc.).

Why 120 DTE?

Let's answer the "why 120 DTE" question first. Refer back to the chart and note the green arrows that show how long the spike in the VIX spanned. It started on Feb 21 and lasted till May 8th. In this case almost 3 months. What we know about VIX spikes is that they occur inversely to SP500 cliff dives. We do not know when they will peak and exactly how long they will last but they don't last very long. And once they start occurring it's almost too late to decide to buy the calls as they rocket so quickly. But we know we want to be ready BEFORE the VIX spike, and we want to have enough to span the event because we do not know when the VIX will peak.

To accomplish that we need multiple months of calls pre-purchased as cheaply as possible. Hence why I purchase them 120 DTE so that my options can span 4 months. One could possibly argue that we can do 90 DTE. As a point of interest, look at the chart and you will note that the peak occurred on March 18 which was exactly the date of expiration for March 2020. However, the highest closing price occurred on March 16. This is coincidental because what if it occurred a week later? Then you would want to be sure we had April calls in play too. The point for me is that we need multiple months to be in play. In 2008, the VIX started spiking in August and the spike lasted till December.

Why 10 deltas?

The reason why I buy 10 deltas comes down to the fact that when the VIX hedge spike it spikes big. I want to own as many call options as possible.. When the VIX spikes typically backwardation is occurring (front months are worth more than back months) and options that were 100% extrinsic value are quickly converted to intrinsic value.

In 2008, the VIX spiked to $96.40 on October 23 but closed at $67.80. That also was not the highest closing day. It closed over $80 the next day and over $79 the day after that. It retreated and spiked again to close over 80 again on November 20th.

Sorry for the ramble but it shows 2 things: 1. The VIX spike can last multiple months and we want to benefit from it as much as possible without being subject to a monthly expiration that occurs in the beginning and miss out on all the benefits of the spike. 2. True VIX spikes will easily envelope the lower deltas and I'll benefit from the greater number of calls because once an option becomes intrinsic the size of the move is constrained as the delta approaches 100.

As an example, I just pulled up the current VIX month and the table below makes the point for $1000 invested in 30 deltas versus 10 deltas:

Code: Select all

 Delta 	 Strike	Price 	Number of Options 
 30.00 	 28.00 	 1.10		  909.09 
 11.00 	 40.00 	 0.40		2,500.00 
So a difference of $12 in strike price but you get 2.75 times more options. In this example, if the VIX closed at $50 on expiration day the 30 deltas would be worth $18,810 whereas the 10 deltas would be worth $24,000 (27.6% more).

Hopefully, this helps answer why 120 DTE and why 10 deltas. As I stated earlier, I want to be mechanical in the execution wherever possible. Currently, it's not that easy since the VIX has remained elevated above historic levels this year so I've made some temporary adjustments.
Awesome, man! Thanks for the detail.
Now it makes perfect sense why you are doing what you are doing.

One more question to confirm my understanding.
Picking 10 deltas means you aren't really going to be protected in 10-20% drop in SP500, right?
You aren't concerned with that as this strategy's purpose is 30%+ crashes, correct?
Perfect Uncertainty
Posts: 16
Joined: Wed Jun 24, 2020 10:20 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Perfect Uncertainty »

djeayzonne wrote: Sun Nov 29, 2020 4:40 pm
Awesome, man! Thanks for the detail.
Now it makes perfect sense why you are doing what you are doing.

One more question to confirm my understanding.
Picking 10 deltas means you aren't really going to be protected in 10-20% drop in SP500, right?
You aren't concerned with that as this strategy's purpose is 30%+ crashes, correct?
That would be a correction versus a Black Swan. One might ask whether TMF is doing its job if we need to hedge a correction with an instrument other than TMF. That is possible but then we should probably be in a different thread for the discussion 😀
Tingting1013
Posts: 721
Joined: Mon Aug 24, 2020 5:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Tingting1013 »

Perfect Uncertainty wrote: Sun Nov 29, 2020 7:21 pm
djeayzonne wrote: Sun Nov 29, 2020 4:40 pm
Awesome, man! Thanks for the detail.
Now it makes perfect sense why you are doing what you are doing.

One more question to confirm my understanding.
Picking 10 deltas means you aren't really going to be protected in 10-20% drop in SP500, right?
You aren't concerned with that as this strategy's purpose is 30%+ crashes, correct?
That would be a correction versus a Black Swan. One might ask whether TMF is doing its job if we need to hedge a correction with an instrument other than TMF. That is possible but then we should probably be in a different thread for the discussion 😀
Is there any scenario where VIX fails to do its job? I.e. the S&P crashes 30%+ and VIX doesn’t move?
cooljack4u
Posts: 6
Joined: Mon Nov 09, 2020 9:51 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by cooljack4u »

perfectuncertainty wrote: Sun Nov 29, 2020 10:13 am
djeayzonne wrote: Sun Nov 29, 2020 3:39 am
Having cleared the commission issue, I am still interested in why you chose the 10 Delta 120 days out. Please elaborate.
Please refer to this spreadsheet as a reference LINK. Note that I added a tab with the VIX chart with some minor annotation.

I prefer to be mechanical in my approach when I execute, but I want a well thought out approach.

So here is a criterion and desired state description:

1. We don't know when a black swan is coming or how long it lasts.
2. We want a preemptive hedge in place for when it does arrive.
3. We want to benefit from tail risk events not be a victim to them.

Since I don't know when a black swan is coming and how long it might last I want to have the hedge in place at all times and I want it to be able to sustain the event. I want to be prepared.

Take a look at the 4-month hedge in the spreadsheet and the VIX Chart I referenced above.

Look at the chart first. The 4 magenta arrows on the left represent the 4 months where each leg of the hedge was purchased. This is described in the summary halfway down under VIX Hedge Preparation (November we pe purchase the 120 DTE 10 delta calls etc.).

Why 120 DTE?

Let's answer the "why 120 DTE" question first. Refer back to the chart and note the green arrows that show how long the spike in the VIX spanned. It started on Feb 21 and lasted till May 8th. In this case almost 3 months. What we know about VIX spikes is that they occur inversely to SP500 cliff dives. We do not know when they will peak and exactly how long they will last but they don't last very long. And once they start occurring it's almost too late to decide to buy the calls as they rocket so quickly. But we know we want to be ready BEFORE the VIX spike, and we want to have enough to span the event because we do not know when the VIX will peak.

To accomplish that we need multiple months of calls pre-purchased as cheaply as possible. Hence why I purchase them 120 DTE so that my options can span 4 months. One could possibly argue that we can do 90 DTE. As a point of interest, look at the chart and you will note that the peak occurred on March 18 which was exactly the date of expiration for March 2020. However, the highest closing price occurred on March 16. This is coincidental because what if it occurred a week later? Then you would want to be sure we had April calls in play too. The point for me is that we need multiple months to be in play. In 2008, the VIX started spiking in August and the spike lasted till December.

Why 10 deltas?

The reason why I buy 10 deltas comes down to the fact that when the VIX hedge spike it spikes big. I want to own as many call options as possible.. When the VIX spikes typically backwardation is occurring (front months are worth more than back months) and options that were 100% extrinsic value are quickly converted to intrinsic value.

In 2008, the VIX spiked to $96.40 on October 23 but closed at $67.80. That also was not the highest closing day. It closed over $80 the next day and over $79 the day after that. It retreated and spiked again to close over 80 again on November 20th.

Sorry for the ramble but it shows 2 things: 1. The VIX spike can last multiple months and we want to benefit from it as much as possible without being subject to a monthly expiration that occurs in the beginning and miss out on all the benefits of the spike. 2. True VIX spikes will easily envelope the lower deltas and I'll benefit from the greater number of calls because once an option becomes intrinsic the size of the move is constrained as the delta approaches 100.

As an example, I just pulled up the current VIX month and the table below makes the point for $1000 invested in 30 deltas versus 10 deltas:

Code: Select all

 Delta 	 Strike	Price 	Number of Options 
 30.00 	 28.00 	 1.10		  909.09 
 11.00 	 40.00 	 0.40		2,500.00 
So a difference of $12 in strike price but you get 2.75 times more options. In this example, if the VIX closed at $50 on expiration day the 30 deltas would be worth $18,810 whereas the 10 deltas would be worth $24,000 (27.6% more).

Hopefully, this helps answer why 120 DTE and why 10 deltas. As I stated earlier, I want to be mechanical in the execution wherever possible. Currently, it's not that easy since the VIX has remained elevated above historic levels this year so I've made some temporary adjustments.
So in your model, the options were in the money on the day of expiry only in 2008 and 2020 crash. Also, if I got it right, you are not suggesting to sell before the expiry but instead let them expire in the money ( since they are cash settled). Did you test if you the money earned by the VIX hedging is reinvested in the HFEA model , what is CAGR compare to not hedging ?
I was looking at VIX in 2001 and 2002 crash, it did not move beyond 45. So, it provides sudden drop and not gradual decay ?
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

Tingting1013 wrote: Sun Nov 29, 2020 7:24 pm
Is there any scenario where VIX fails to do its job? I.e. the S&P crashes 30%+ and VIX doesn’t move?
The VIX is a direct calculation based on the S&P - it's not possible for it not to move if the S&P moves.
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

cooljack4u wrote: Sun Nov 29, 2020 8:14 pm So in your model, the options were in the money on the day of expiry only in 2008 and 2020 crash. Also, if I got it right, you are not suggesting to sell before the expiry but instead let them expire in the money ( since they are cash settled).
When to sell is a function of how much offset we have in a black swan event. The VIX reverts quickly - as volatility subsides. So HFEA could settle quietly at a much lower price and quell volatility. My preference is to obtain 10 delta calls at a strike price of 45 or below. I'm looking to exit at 65 or above - so at that point it’s a judgment call. For me, I don't sell on the first peak over 65, and if it's close enough to expiration (3 days or less), I'd probably let it expire as expiration day is often volatile too.
Did you test if you the money earned by the VIX hedging is reinvested in the HFEA model , what is CAGR compare to not hedging ?
Refer to the spreadsheet I provided previously to see the impact on HFEA hedged versus unhedged.
I was looking at VIX in 2001 and 2002 crash, it did not move beyond 45. So, it provides sudden drop and not gradual decay ?
Yes, 2001-2002 was not a black swan event and the market declined gradually, so TMF should have partially hedged the decline in a cyclical decline (SP500 declined 40% and bonds rose 25% at that time). Again the VIX hedge is preemptive. A decline such as 2001-2002 did not require preemptive insurance.
cooljack4u
Posts: 6
Joined: Mon Nov 09, 2020 9:51 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by cooljack4u »

perfectuncertainty wrote: Sun Nov 29, 2020 9:49 pm
cooljack4u wrote: Sun Nov 29, 2020 8:14 pm So in your model, the options were in the money on the day of expiry only in 2008 and 2020 crash. Also, if I got it right, you are not suggesting to sell before the expiry but instead let them expire in the money ( since they are cash settled).
When to sell is a function of how much offset we have in a black swan event. The VIX reverts quickly - as volatility subsides. So HFEA could settle quietly at a much lower price and quell volatility. My preference is to obtain 10 delta calls at a strike price of 45 or below. I'm looking to exit at 65 or above - so at that point it’s a judgment call. For me, I don't sell on the first peak over 65, and if it's close enough to expiration (3 days or less), I'd probably let it expire as expiration day is often volatile too.
Did you test if you the money earned by the VIX hedging is reinvested in the HFEA model , what is CAGR compare to not hedging ?
Refer to the spreadsheet I provided previously to see the impact on HFEA hedged versus unhedged.
I was looking at VIX in 2001 and 2002 crash, it did not move beyond 45. So, it provides sudden drop and not gradual decay ?
Yes, 2001-2002 was not a black swan event and the market declined gradually, so TMF should have partially hedged the decline in a cyclical decline (SP500 declined 40% and bonds rose 25% at that time). Again the VIX hedge is preemptive. A decline such as 2001-2002 did not require preemptive insurance.
My point is that if you had reinvested the VIX Hedge profit back into HFEA ( that is pumping back 580K into the model, I am sure it would give handsome results. Similarly, if the 2008 Hedge income is reinvested back in HFEA overall CAGR will be far better, isnt it ?
cooljack4u
Posts: 6
Joined: Mon Nov 09, 2020 9:51 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by cooljack4u »

perfectuncertainty wrote: Sun Nov 29, 2020 9:49 pm
cooljack4u wrote: Sun Nov 29, 2020 8:14 pm So in your model, the options were in the money on the day of expiry only in 2008 and 2020 crash. Also, if I got it right, you are not suggesting to sell before the expiry but instead let them expire in the money ( since they are cash settled).
When to sell is a function of how much offset we have in a black swan event. The VIX reverts quickly - as volatility subsides. So HFEA could settle quietly at a much lower price and quell volatility. My preference is to obtain 10 delta calls at a strike price of 45 or below. I'm looking to exit at 65 or above - so at that point it’s a judgment call. For me, I don't sell on the first peak over 65, and if it's close enough to expiration (3 days or less), I'd probably let it expire as expiration day is often volatile too.
Did you test if you the money earned by the VIX hedging is reinvested in the HFEA model , what is CAGR compare to not hedging ?
Refer to the spreadsheet I provided previously to see the impact on HFEA hedged versus unhedged.
I was looking at VIX in 2001 and 2002 crash, it did not move beyond 45. So, it provides sudden drop and not gradual decay ?
Yes, 2001-2002 was not a black swan event and the market declined gradually, so TMF should have partially hedged the decline in a cyclical decline (SP500 declined 40% and bonds rose 25% at that time). Again the VIX hedge is preemptive. A decline such as 2001-2002 did not require preemptive insurance.
Also, here is my idea to pay for the hedge, write monthly OTM calls that pays premium of 0.6 percent. Now, we will write these call only in the month where last month closed positive. Since UPRO startetd, there are 91 months/139 months where you would be able to collect the premium. These months, the max move positive move was 22 percent. If we write monthly 15 to 20 Percent OTM calls, we should be able to collect enough premium that will pay for premium plus any extra profit that will be left ( if in case the calls closes ITM). This will require little bit of deeper analysis of UPRO trend in terms of past few months returns to decide which calls to write (15 or 20 percent OTM )
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

cooljack4u wrote: Sun Nov 29, 2020 10:01 pm
My point is that if you had reinvested the VIX Hedge profit back into HFEA ( that is pumping back 580K into the model, I am sure it would give handsome results. Similarly, if the 2008 Hedge income is reinvested back in HFEA overall CAGR will be far better, isnt it ?
no doubt
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

cooljack4u wrote: Sun Nov 29, 2020 10:25 pm Also, here is my idea to pay for the hedge, write monthly OTM calls that pays premium of 0.6 percent. Now, we will write these call only in the month where last month closed positive. Since UPRO startetd, there are 91 months/139 months where you would be able to collect the premium. These months, the max move positive move was 22 percent. If we write monthly 15 to 20 Percent OTM calls, we should be able to collect enough premium that will pay for premium plus any extra profit that will be left ( if in case the calls closes ITM). This will require little bit of deeper analysis of UPRO trend in terms of past few months returns to decide which calls to write (15 or 20 percent OTM )
Jan 21 UPRO 90 strike will get you 0.64% as of Friday's close.
User avatar
Uncorrelated
Posts: 1235
Joined: Sun Oct 13, 2019 3:16 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Uncorrelated »

perfectuncertainty wrote: Sun Nov 29, 2020 8:36 pm
Tingting1013 wrote: Sun Nov 29, 2020 7:24 pm
Is there any scenario where VIX fails to do its job? I.e. the S&P crashes 30%+ and VIX doesn’t move?
The VIX is a direct calculation based on the S&P - it's not possible for it not to move if the S&P moves.
The VIX is based on the options implied volatility and market risk aversion. If the market drops but options-implies volatility stays the same, the VIX won't move. In practice there is a high negative correlation between VIX and the stock markets, but that is not a guarantee.

The negative correlation is actually problematic in a trading strategy. It means that buying vix (or options) comes with significant exposure to market beta. However, in the absence of predictability of returns you'd want to keep a constant exposure to beta (no, you cannot predict returns). Also, since VIX is significantly correlated with negative beta, purchasing VIX calls is roughly equivalent with reducing the equity exposure. But if it's your goal to time the market and reduce equity exposure in bad times, reducing leverage (i.e. tossing UPRO) is much cheaper than option plays.


I have been doing a lot of research on volatility trading strategies. The academic literature suggests there is a very strong positive premium attached with bearing downside risk. Since you are hedging away the downside risk, you can expect to pay heavily for that insurance. It's really unclear to me what the arguments for this strategy are in the first place. The only thing that's clear here is that the strategy is heavily overfitted.
cooljack4u
Posts: 6
Joined: Mon Nov 09, 2020 9:51 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by cooljack4u »

Uncorrelated wrote: Mon Nov 30, 2020 9:00 am
perfectuncertainty wrote: Sun Nov 29, 2020 8:36 pm
Tingting1013 wrote: Sun Nov 29, 2020 7:24 pm
Is there any scenario where VIX fails to do its job? I.e. the S&P crashes 30%+ and VIX doesn’t move?
The VIX is a direct calculation based on the S&P - it's not possible for it not to move if the S&P moves.
The VIX is based on the options implied volatility and market risk aversion. If the market drops but options-implies volatility stays the same, the VIX won't move. In practice there is a high negative correlation between VIX and the stock markets, but that is not a guarantee.

The negative correlation is actually problematic in a trading strategy. It means that buying vix (or options) comes with significant exposure to market beta. However, in the absence of predictability of returns you'd want to keep a constant exposure to beta (no, you cannot predict returns). Also, since VIX is significantly correlated with negative beta, purchasing VIX calls is roughly equivalent with reducing the equity exposure. But if it's your goal to time the market and reduce equity exposure in bad times, reducing leverage (i.e. tossing UPRO) is much cheaper than option plays.


I have been doing a lot of research on volatility trading strategies. The academic literature suggests there is a very strong positive premium attached with bearing downside risk. Since you are hedging away the downside risk, you can expect to pay heavily for that insurance. It's really unclear to me what the arguments for this strategy are in the first place. The only thing that's clear here is that the strategy is heavily overfitted.
What do you mean by "But if it's your goal to time the market and reduce equity exposure in bad times, reducing leverage (i.e. tossing UPRO) is much cheaper than option plays"

How will you come to know when to reduce UPRO ? With strategy VIX mentioned by pefectuncertainty, you are doing that on a regular basis and hoping to cash it out every 10 years or so. I understand this is a slow and painful but what other alternative is there to not go through emotion of loosing 70 percent ( of UPRO that happened in March) and still be in the market to ride the next high tide
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

Uncorrelated wrote: Mon Nov 30, 2020 9:00 am
perfectuncertainty wrote: Sun Nov 29, 2020 8:36 pm
Tingting1013 wrote: Sun Nov 29, 2020 7:24 pm
Is there any scenario where VIX fails to do its job? I.e. the S&P crashes 30%+ and VIX doesn’t move?
The VIX is a direct calculation based on the S&P - it's not possible for it not to move if the S&P moves.
The VIX is based on the options implied volatility and market risk aversion. If the market drops but options-implies volatility stays the same, the VIX won't move. In practice there is a high negative correlation between VIX and the stock markets, but that is not a guarantee.

The negative correlation is actually problematic in a trading strategy. It means that buying vix (or options) comes with significant exposure to market beta. However, in the absence of predictability of returns you'd want to keep a constant exposure to beta (no, you cannot predict returns). Also, since VIX is significantly correlated with negative beta, purchasing VIX calls is roughly equivalent with reducing the equity exposure. But if it's your goal to time the market and reduce equity exposure in bad times, reducing leverage (i.e. tossing UPRO) is much cheaper than option plays.


I have been doing a lot of research on volatility trading strategies. The academic literature suggests there is a very strong positive premium attached with bearing downside risk. Since you are hedging away the downside risk, you can expect to pay heavily for that insurance. It's really unclear to me what the arguments for this strategy are in the first place. The only thing that's clear here is that the strategy is heavily overfitted.
Thanks for your response. I agree with you and pretty much described such in my next post where I addressed the 2001 - 2002 decline. That being said the question posed to me was "Is there any scenario where VIX fails to do its job? I.e. the S&P crashes 30%+ and VIX doesn’t move?" I should have been more clear but I stand by the spirit of my answer that there is no way that the SP500 crashes 30% and the VIX doesn’t move in response.
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

cooljack4u wrote: Mon Nov 30, 2020 12:04 pm
Uncorrelated wrote: Mon Nov 30, 2020 9:00 am
perfectuncertainty wrote: Sun Nov 29, 2020 8:36 pm
Tingting1013 wrote: Sun Nov 29, 2020 7:24 pm
Is there any scenario where VIX fails to do its job? I.e. the S&P crashes 30%+ and VIX doesn’t move?
The VIX is a direct calculation based on the S&P - it's not possible for it not to move if the S&P moves.
The VIX is based on the options implied volatility and market risk aversion. If the market drops but options-implies volatility stays the same, the VIX won't move. In practice there is a high negative correlation between VIX and the stock markets, but that is not a guarantee.

The negative correlation is actually problematic in a trading strategy. It means that buying vix (or options) comes with significant exposure to market beta. However, in the absence of predictability of returns you'd want to keep a constant exposure to beta (no, you cannot predict returns). Also, since VIX is significantly correlated with negative beta, purchasing VIX calls is roughly equivalent with reducing the equity exposure. But if it's your goal to time the market and reduce equity exposure in bad times, reducing leverage (i.e. tossing UPRO) is much cheaper than option plays.


I have been doing a lot of research on volatility trading strategies. The academic literature suggests there is a very strong positive premium attached with bearing downside risk. Since you are hedging away the downside risk, you can expect to pay heavily for that insurance. It's really unclear to me what the arguments for this strategy are in the first place. The only thing that's clear here is that the strategy is heavily overfitted.
What do you mean by "But if it's your goal to time the market and reduce equity exposure in bad times, reducing leverage (i.e. tossing UPRO) is much cheaper than option plays"

How will you come to know when to reduce UPRO ? With strategy VIX mentioned by pefectuncertainty, you are doing that on a regular basis and hoping to cash it out every 10 years or so. I understand this is a slow and painful but what other alternative is there to not go through emotion of loosing 70 percent ( of UPRO that happened in March) and still be in the market to ride the next high tide
As I stated, I prefer to be mechanical and remove emotion from my trading. However, if you know specifically when to reduce UPRO, please let me know and I will guarantee I'll make an absolute killing with options if your timing is correct. I don't know when the market is going to crash.

I do agree there are certainly times when probability can point to strategies to use versus other strategies.
User avatar
willthrill81
Posts: 22728
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by willthrill81 »

I find the suggestion to use VIX rather than TMF very interesting. Using it instead of TMF in a target volatility approach, which is what I'm personally using in my own 'excellent adventure', would have been very fruitful.

Below are the results of a 25% target volatility approach with 70% in UPRO and 30% in VIX.

Image

I'll have to mull this one over a bit. Thoughts?
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings
Jags4186
Posts: 5372
Joined: Wed Jun 18, 2014 7:12 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Jags4186 »

willthrill81 wrote: Mon Nov 30, 2020 1:53 pm I find the suggestion to use VIX rather than TMF very interesting. Using it instead of TMF in a target volatility approach, which is what I'm personally using in my own 'excellent adventure', would have been very fruitful.

Below are the results of a 25% target volatility approach with 70% in UPRO and 30% in VIX.

Image

I'll have to mull this one over a bit. Thoughts?
I think the main question is what instrument do you intend to use to capture ^VIX. I ran the same assumptions which you used and used ticker XVZ. It only goes back to January 2012 so let's use those numbers:

^VIX Index results

https://www.portfoliovisualizer.com/tes ... tion2_1=70

vs. XVZ results

https://www.portfoliovisualizer.com/tes ... tion2_1=30
langlands
Posts: 758
Joined: Wed Apr 03, 2019 10:05 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by langlands »

willthrill81 wrote: Mon Nov 30, 2020 1:53 pm I find the suggestion to use VIX rather than TMF very interesting. Using it instead of TMF in a target volatility approach, which is what I'm personally using in my own 'excellent adventure', would have been very fruitful.

Below are the results of a 25% target volatility approach with 70% in UPRO and 30% in VIX.

Image

I'll have to mull this one over a bit. Thoughts?
It's not possible to actually invest directly in the VIX index. You'd have to buy futures or invest in VXX and suffer the roll cost (VIX futures are almost always in contango during bull markets).
User avatar
willthrill81
Posts: 22728
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by willthrill81 »

Jags4186 wrote: Mon Nov 30, 2020 2:22 pm
willthrill81 wrote: Mon Nov 30, 2020 1:53 pm I find the suggestion to use VIX rather than TMF very interesting. Using it instead of TMF in a target volatility approach, which is what I'm personally using in my own 'excellent adventure', would have been very fruitful.

Below are the results of a 25% target volatility approach with 70% in UPRO and 30% in VIX.

Image

I'll have to mull this one over a bit. Thoughts?
I think the main question is what instrument do you intend to use to capture ^VIX.
langlands wrote: Mon Nov 30, 2020 2:25 pm It's not possible to actually invest directly in the VIX index. You'd have to buy futures or invest in VXX and suffer the roll cost (VIX futures are almost always in contango during bull markets).
Thanks. I guess that kills that strategy. :x
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings
User avatar
Uncorrelated
Posts: 1235
Joined: Sun Oct 13, 2019 3:16 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Uncorrelated »

cooljack4u wrote: Mon Nov 30, 2020 12:04 pm
Uncorrelated wrote: Mon Nov 30, 2020 9:00 am
perfectuncertainty wrote: Sun Nov 29, 2020 8:36 pm
Tingting1013 wrote: Sun Nov 29, 2020 7:24 pm
Is there any scenario where VIX fails to do its job? I.e. the S&P crashes 30%+ and VIX doesn’t move?
The VIX is a direct calculation based on the S&P - it's not possible for it not to move if the S&P moves.
The VIX is based on the options implied volatility and market risk aversion. If the market drops but options-implies volatility stays the same, the VIX won't move. In practice there is a high negative correlation between VIX and the stock markets, but that is not a guarantee.

The negative correlation is actually problematic in a trading strategy. It means that buying vix (or options) comes with significant exposure to market beta. However, in the absence of predictability of returns you'd want to keep a constant exposure to beta (no, you cannot predict returns). Also, since VIX is significantly correlated with negative beta, purchasing VIX calls is roughly equivalent with reducing the equity exposure. But if it's your goal to time the market and reduce equity exposure in bad times, reducing leverage (i.e. tossing UPRO) is much cheaper than option plays.


I have been doing a lot of research on volatility trading strategies. The academic literature suggests there is a very strong positive premium attached with bearing downside risk. Since you are hedging away the downside risk, you can expect to pay heavily for that insurance. It's really unclear to me what the arguments for this strategy are in the first place. The only thing that's clear here is that the strategy is heavily overfitted.
What do you mean by "But if it's your goal to time the market and reduce equity exposure in bad times, reducing leverage (i.e. tossing UPRO) is much cheaper than option plays"

How will you come to know when to reduce UPRO ? With strategy VIX mentioned by pefectuncertainty, you are doing that on a regular basis and hoping to cash it out every 10 years or so. I understand this is a slow and painful but what other alternative is there to not go through emotion of loosing 70 percent ( of UPRO that happened in March) and still be in the market to ride the next high tide
I was thinking about market timing possibilities. perfectuncertainty has mentioned that he made some temporarily adjustments due to heightened VIX levels. This strongly hints towards market timing.

The cashing out is another aspect that strongly hints at market timing. Unless there are very specific time-varying expectations, it is optimal to hold a constant (percentage of total portfolio) exposure to VIX. Never cash out, just keep rolling your positions.

I personally don't believe (read: have not seen any evidence indicating) that market timing in this way can work. Of course it is possible to trade VIX without market timing if you just hold a constant exposure.

What you shouldn't forget is that these instruments are priced in an efficient market, and the efficient market is mostly made up of hedge funds and big banks that can leverage much cheaper than you. Even if hedge funds can make money with a VIX strategy, by the time you factor in the higher expense ratio of UPRO and massive information deficient, it's unlikely you will be able to capture much of a premium. Instead of buying a triple-leveraged ETF and then purchasing complicated derivatives to hedge that risk, it sounds like a much better idea to avoid the triple leveraged ETF in the first place. Certainly, the evidence I have seen is to weak to take VIX strategies into serious consideration.
langlands
Posts: 758
Joined: Wed Apr 03, 2019 10:05 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by langlands »

Uncorrelated wrote: Mon Nov 30, 2020 5:39 pm
cooljack4u wrote: Mon Nov 30, 2020 12:04 pm
Uncorrelated wrote: Mon Nov 30, 2020 9:00 am
perfectuncertainty wrote: Sun Nov 29, 2020 8:36 pm
Tingting1013 wrote: Sun Nov 29, 2020 7:24 pm
Is there any scenario where VIX fails to do its job? I.e. the S&P crashes 30%+ and VIX doesn’t move?
The VIX is a direct calculation based on the S&P - it's not possible for it not to move if the S&P moves.
The VIX is based on the options implied volatility and market risk aversion. If the market drops but options-implies volatility stays the same, the VIX won't move. In practice there is a high negative correlation between VIX and the stock markets, but that is not a guarantee.

The negative correlation is actually problematic in a trading strategy. It means that buying vix (or options) comes with significant exposure to market beta. However, in the absence of predictability of returns you'd want to keep a constant exposure to beta (no, you cannot predict returns). Also, since VIX is significantly correlated with negative beta, purchasing VIX calls is roughly equivalent with reducing the equity exposure. But if it's your goal to time the market and reduce equity exposure in bad times, reducing leverage (i.e. tossing UPRO) is much cheaper than option plays.


I have been doing a lot of research on volatility trading strategies. The academic literature suggests there is a very strong positive premium attached with bearing downside risk. Since you are hedging away the downside risk, you can expect to pay heavily for that insurance. It's really unclear to me what the arguments for this strategy are in the first place. The only thing that's clear here is that the strategy is heavily overfitted.
What do you mean by "But if it's your goal to time the market and reduce equity exposure in bad times, reducing leverage (i.e. tossing UPRO) is much cheaper than option plays"

How will you come to know when to reduce UPRO ? With strategy VIX mentioned by pefectuncertainty, you are doing that on a regular basis and hoping to cash it out every 10 years or so. I understand this is a slow and painful but what other alternative is there to not go through emotion of loosing 70 percent ( of UPRO that happened in March) and still be in the market to ride the next high tide
I was thinking about market timing possibilities. perfectuncertainty has mentioned that he made some temporarily adjustments due to heightened VIX levels. This strongly hints towards market timing.

The cashing out is another aspect that strongly hints at market timing. Unless there are very specific time-varying expectations, it is optimal to hold a constant (percentage of total portfolio) exposure to VIX. Never cash out, just keep rolling your positions.

I personally don't believe (read: have not seen any evidence indicating) that market timing in this way can work. Of course it is possible to trade VIX without market timing if you just hold a constant exposure.

What you shouldn't forget is that these instruments are priced in an efficient market, and the efficient market is mostly made up of hedge funds and big banks that can leverage much cheaper than you. Even if hedge funds can make money with a VIX strategy, by the time you factor in the higher expense ratio of UPRO and massive information deficient, it's unlikely you will be able to capture much of a premium. Instead of buying a triple-leveraged ETF and then purchasing complicated derivatives to hedge that risk, it sounds like a much better idea to avoid the triple leveraged ETF in the first place. Certainly, the evidence I have seen is to weak to take VIX strategies into serious consideration.
I agree that trading VIX-related products successfully requires a level of sophistication higher than that of leveraged ETFs or SPY futures or vanilla options.

However, I don't see a problem with using VIX as an input to decide how much leverage to use. Say you are only deciding between two assets, SPY or cash. Merton says your allocation should be proportional to mu/sigma^2 where mu is the expected return on SPY and sigma is volatility of SPY. What should you do if volatility doubles? If you don't think the risk premium changes, then you should reduce your allocation by a factor of 4. If you think risk premium increases by a factor of 2, then you should reduce your allocation by a factor of 2. If you think risk premium increases by a factor of 4, then you shouldn't change anything.

For people who care about optimal portfolio allocation, I never really understood the rationale for holding through a crash. It seems the main Boglehead reasons are purely psychological (basically, people never end up buying back in). But ignoring the psychological reasons, it seems pretty clear that your equity allocation should decrease if volatility goes up by a lot. Certainly if you are leveraged to a significant extent, not deleveraging during periods of high volatility seems unwise.
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

Uncorrelated wrote: Mon Nov 30, 2020 5:39 pm
I was thinking about market timing possibilities. perfectuncertainty has mentioned that he made some temporarily adjustments due to heightened VIX levels. This strongly hints towards market timing.
How sure are you?

The temporary adjustments have nothing to do with timing. They have to do with pricing and probability. When the VIX has remained high (as it has done this year) and you can only get 85 strike prices as 10 deltas 120 DTE and you know that only twice in history has the VIX traded at those levels and call options are unlikely to convert IV to intrinsic with much overlap, if any, then you make adjustments. Like looking at the front month to see if you can pick them up without the built-in theta, for instance.

Not timing - valuation and probability.
The cashing out is another aspect that strongly hints at market timing. Unless there are very specific time-varying expectations, it is optimal to hold a constant (percentage of total portfolio) exposure to VIX. Never cash out, just keep rolling your positions.
You think? So when I cashed out in March because my VIX call options were through the roof and we know the VIX reverts quite quickly - that's not timing. That's profit-taking - lots of it. When we have serious backwardation and a massive spike on the front month and our call options are deep ITM and have 10s of dollars of intrinsic value and we cash - that's not timing. Or would you sit and just hold the VIX instruments you have (because academically that might be timing)? C'mon, you are incredibly intelligent (super high IQ probably) - maybe you can learn from some others?

I personally don't believe (read: have not seen any evidence indicating) that market timing in this way can work. Of course it is possible to trade VIX without market timing if you just hold a constant exposure.

What you shouldn't forget is that these instruments are priced in an efficient market, and the efficient market is mostly made up of hedge funds and big banks that can leverage much cheaper than you. Even if hedge funds can make money with a VIX strategy, by the time you factor in the higher expense ratio of UPRO and massive information deficient, it's unlikely you will be able to capture much of a premium. Instead of buying a triple-leveraged ETF and then purchasing complicated derivatives to hedge that risk, it sounds like a much better idea to avoid the triple leveraged ETF in the first place. Certainly, the evidence I have seen is to weak to take VIX strategies into serious consideration.
How do you trade VIX without cashing it in?
Care to demonstrate your real account with offsets or hedges that rose during March of this year?
User avatar
Uncorrelated
Posts: 1235
Joined: Sun Oct 13, 2019 3:16 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Uncorrelated »

langlands wrote: Mon Nov 30, 2020 6:43 pm
Uncorrelated wrote: Mon Nov 30, 2020 5:39 pm snip
I agree that trading VIX-related products successfully requires a level of sophistication higher than that of leveraged ETFs or SPY futures or vanilla options.

However, I don't see a problem with using VIX as an input to decide how much leverage to use. Say you are only deciding between two assets, SPY or cash. Merton says your allocation should be proportional to mu/sigma^2 where mu is the expected return on SPY and sigma is volatility of SPY. What should you do if volatility doubles? If you don't think the risk premium changes, then you should reduce your allocation by a factor of 4. If you think risk premium increases by a factor of 2, then you should reduce your allocation by a factor of 2. If you think risk premium increases by a factor of 4, then you shouldn't change anything.

For people who care about optimal portfolio allocation, I never really understood the rationale for holding through a crash. It seems the main Boglehead reasons are purely psychological (basically, people never end up buying back in). But ignoring the psychological reasons, it seems pretty clear that your equity allocation should decrease if volatility goes up by a lot. Certainly if you are leveraged to a significant extent, not deleveraging during periods of high volatility seems unwise.
This is exactly how the theory works, I have referred to this theory many times in this forum. It's so convincing that I spent a considerable amount of time analyzing this market timing strategy. But I was simple unable to get it to work reliable.

The first problem is that VIX actually isn't that good of a predictor of volatility. I'm not sure about the exact reasons for this, but I do know that performing Merton's portfolio problem with VIX as the expected volatility results in lower out-of-sample CER (certainty equivalent return) than a constant allocation.

Thankfully, I've found it to be relatively trivial to create a better volatility estimate with a few regression models. This results in a marked increase in CER, at least on my dataset (1932-2020) with daily rebalancing, so that's great. Problem is: volatility changes fast. Should we rebalance daily? Monthly? Somewhere in between? I used reinforcement learning to calculate the optimal trading policy given volatility forecasts over various time horizons with the objective to maximize expected utility after trading costs. My best model was a slight improvement over a constant allocation.

My reinforcement learning process was only half out of sample, so we should have some skepticism there. But a better argument is that the economic significance just wasn't very large: a lifecycle model or value tilt results in far higher CER gains. If you already have lifecycle and factor tilting, does volatility timing still results in CER gains? I don't know, but based on my experiments I don't expect the gains to be all that large.


It's a good idea, well supported in theory and empirical data, I just wasn't able to get a complete system to work. There is a fair amount of literature to claims this will work, but as far as I'm aware of all literature except this paper which I can't read suffers from look-ahead bias and questionable statistical methods.
User avatar
Uncorrelated
Posts: 1235
Joined: Sun Oct 13, 2019 3:16 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Uncorrelated »

perfectuncertainty wrote: Mon Nov 30, 2020 11:18 pm
Uncorrelated wrote: Mon Nov 30, 2020 5:39 pm
I was thinking about market timing possibilities. perfectuncertainty has mentioned that he made some temporarily adjustments due to heightened VIX levels. This strongly hints towards market timing.
How sure are you?

The temporary adjustments have nothing to do with timing. They have to do with pricing and probability. When the VIX has remained high (as it has done this year) and you can only get 85 strike prices as 10 deltas 120 DTE and you know that only twice in history has the VIX traded at those levels and call options are unlikely to convert IV to intrinsic with much overlap, if any, then you make adjustments. Like looking at the front month to see if you can pick them up without the built-in theta, for instance.

Not timing - valuation and probability.
Looking at valuations is the definition of timing.

Let's take the equity risk premium. The equity risk premium is more or less constant over time. With a few more assumptions, the optimal approach is to have a constant allocation to stocks.

It's still unclear to me what the arguments for VIX are, but a possible reason is that holding long VIX is rewarded with a risk premium. In this case again, it optimal to hold a constant exposure to VIX.

Assuming markets are moderately efficient, you cannot make money by out-predicting the VIX. Your only hope at profits is to capture some risk premium. If you are willing to purchase VIX (i.e. capture some risk premium) at time A but not at time B, that is almost equivalent with stating markets are grossly inefficient. In an efficient market, if the VIX looks unattractive, it means that many people want to hedge because the probability of a negative event is high. It sounds like you don't want to hedge precisely when the probability of bad events is high. That doesn't make any sense.

There is one more argument why your strategy would be market timing. Remember that I said that VIX has a negative correlation with market beta, and that the optimal allocation to market beta is a constant? If you're usually holding VIX, but aren't holding VIX now, that means your exposure to market beta is higher than it is normally. Even if what you're doing should not be considered timing VIX, you're surely timing market beta.


Anyway, I'm not saying that timing the VIX doesn't work. I'm just saying that you're timing. Timing usually only works if the market is grossly inefficient. So perhaps you can show the market is grossly inefficient?
The cashing out is another aspect that strongly hints at market timing. Unless there are very specific time-varying expectations, it is optimal to hold a constant (percentage of total portfolio) exposure to VIX. Never cash out, just keep rolling your positions.
You think? So when I cashed out in March because my VIX call options were through the roof and we know the VIX reverts quite quickly - that's not timing. That's profit-taking - lots of it. When we have serious backwardation and a massive spike on the front month and our call options are deep ITM and have 10s of dollars of intrinsic value and we cash - that's not timing. Or would you sit and just hold the VIX instruments you have (because academically that might be timing)? C'mon, you are incredibly intelligent (super high IQ probably) - maybe you can learn from some others?
What you're really saying is that you can predict the VIX with greater accuracy than the market.

Would love to see any research indicating that this is possible. If you can predict the VIX better than the market, that would pretty clearly indicate that the markets are inefficient. Where are all the active managers making money on VIX futures? If you are making money, who is on the other side of that trade?

If you perform a call or put strategy (for example: put with strike 10% below current price) and the price moves such that the put no longer has a strike 10% below current price, you should probably rebalance if you want to avoid market timing and anchoring. I suppose you could call that "cashing out", but I call it rebalancing. I explained my reasoning in this topic I opened last week. This might sound contradictory (after all, this way you'll never capitalize on the put...), but is the only strategy consistent with some very fundamental mathematical theories. Unless you believe you can time the markets of course, then the theory goes out of the window.

I personally don't believe (read: have not seen any evidence indicating) that market timing in this way can work. Of course it is possible to trade VIX without market timing if you just hold a constant exposure.

What you shouldn't forget is that these instruments are priced in an efficient market, and the efficient market is mostly made up of hedge funds and big banks that can leverage much cheaper than you. Even if hedge funds can make money with a VIX strategy, by the time you factor in the higher expense ratio of UPRO and massive information deficient, it's unlikely you will be able to capture much of a premium. Instead of buying a triple-leveraged ETF and then purchasing complicated derivatives to hedge that risk, it sounds like a much better idea to avoid the triple leveraged ETF in the first place. Certainly, the evidence I have seen is to weak to take VIX strategies into serious consideration.
How do you trade VIX without cashing it in?
Care to demonstrate your real account with offsets or hedges that rose during March of this year?
I would trade VIX the same way you would trade market beta: by holding a constant allocation to the corresponding risk factors and rebalancing regularly. If you don't believe there are risk factors, well, that would indicate market timing or arbitrage, and this ain't arbitrage. :happy

My real account doesn't trade options or volatility because I haven't been able to uncover any strong evidence indicating this would result in higher CER. All academic evidence points in the opposite direction that you are trading: long volatility is a net loss position and short volatility is a net win position. This is not the result of market inefficiency, but a premium for unhedgeable risk that occurs as a byproduct of delta-hedging options.
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

Uncorrelated wrote: Tue Dec 01, 2020 4:42 am
perfectuncertainty wrote: Mon Nov 30, 2020 11:18 pm
Uncorrelated wrote: Mon Nov 30, 2020 5:39 pm
I was thinking about market timing possibilities. perfectuncertainty has mentioned that he made some temporarily adjustments due to heightened VIX levels. This strongly hints towards market timing.
How sure are you?

The temporary adjustments have nothing to do with timing. They have to do with pricing and probability. When the VIX has remained high (as it has done this year) and you can only get 85 strike prices as 10 deltas 120 DTE and you know that only twice in history has the VIX traded at those levels and call options are unlikely to convert IV to intrinsic with much overlap, if any, then you make adjustments. Like looking at the front month to see if you can pick them up without the built-in theta, for instance.

Not timing - valuation and probability.
Looking at valuations is the definition of timing.
You advocate for rebalancing. Why rebalance? Because valuation in your portfolio is skewed? Makes sense - but according to you, that would be timing. If you rebalance to align with probability - that's precisely why I don't buy VIX options that will not have value in the event they have little to no probability of having multiples of intrinsic value if the VIX retests its highs.
Let's take the equity risk premium. The equity risk premium is more or less constant over time. With a few more assumptions, the optimal approach is to have a constant allocation to stocks.

It's still unclear to me what the arguments for VIX are, but a possible reason is that holding long VIX is rewarded with a risk premium. In this case again, it optimal to hold a constant exposure to VIX.

Assuming markets are moderately efficient, you cannot make money by out-predicting the VIX. Your only hope at profits is to capture some risk premium. If you are willing to purchase VIX (i.e. capture some risk premium) at time A but not at time B, that is almost equivalent with stating markets are grossly inefficient. In an efficient market, if the VIX looks unattractive, it means that many people want to hedge because the probability of a negative event is high. It sounds like you don't want to hedge precisely when the probability of bad events is high. That doesn't make any sense.

There is one more argument why your strategy would be market timing. Remember that I said that VIX has a negative correlation with market beta, and that the optimal allocation to market beta is a constant? If you're usually holding VIX, but aren't holding VIX now, that means your exposure to market beta is higher than it is normally. Even if what you're doing should not be considered timing VIX, you're surely timing market beta.


Anyway, I'm not saying that timing the VIX doesn't work. I'm just saying that you're timing. Timing usually only works if the market is grossly inefficient. So perhaps you can show the market is grossly inefficient?
The market is not always efficient - depending on your definition of the market. I know your book says it is, but tell me why massive backwardation is currently occurring in Tesla call options right now? Market efficiency? I'd love to hear your answer to this question.
The cashing out is another aspect that strongly hints at market timing. Unless there are very specific time-varying expectations, it is optimal to hold a constant (percentage of total portfolio) exposure to VIX. Never cash out, just keep rolling your positions.
You think? So when I cashed out in March because my VIX call options were through the roof and we know the VIX reverts quite quickly - that's not timing. That's profit-taking - lots of it. When we have serious backwardation and a massive spike on the front month and our call options are deep ITM and have 10s of dollars of intrinsic value and we cash - that's not timing. Or would you sit and just hold the VIX instruments you have (because academically that might be timing)? C'mon, you are incredibly intelligent (super high IQ probably) - maybe you can learn from some others?
What you're really saying is that you can predict the VIX with greater accuracy than the market.

Would love to see any research indicating that this is possible. If you can predict the VIX better than the market, that would pretty clearly indicate that the markets are inefficient. Where are all the active managers making money on VIX futures? If you are making money, who is on the other side of that trade?

If you perform a call or put strategy (for example: put with strike 10% below current price) and the price moves such that the put no longer has a strike 10% below current price, you should probably rebalance if you want to avoid market timing and anchoring. I suppose you could call that "cashing out", but I call it rebalancing. I explained my reasoning in this topic I opened last week. This might sound contradictory (after all, this way you'll never capitalize on the put...), but is the only strategy consistent with some very fundamental mathematical theories. Unless you believe you can time the markets of course, then the theory goes out of the window.
Nope. Not making any predictions. I do not know when the VIX is going to spike (stated such multiple times). I do know that reversion of the VIX occurs quite quickly and that when I have multiple tens of thousands of options that are deep ITM and the VIX is over 65 I will gladly cash out every time.

Or do you dispute VIX reversion?

I personally don't believe (read: have not seen any evidence indicating) that market timing in this way can work. Of course it is possible to trade VIX without market timing if you just hold a constant exposure.

What you shouldn't forget is that these instruments are priced in an efficient market, and the efficient market is mostly made up of hedge funds and big banks that can leverage much cheaper than you. Even if hedge funds can make money with a VIX strategy, by the time you factor in the higher expense ratio of UPRO and massive information deficient, it's unlikely you will be able to capture much of a premium. Instead of buying a triple-leveraged ETF and then purchasing complicated derivatives to hedge that risk, it sounds like a much better idea to avoid the triple leveraged ETF in the first place. Certainly, the evidence I have seen is to weak to take VIX strategies into serious consideration.
How do you trade VIX without cashing it in?
Care to demonstrate your real account with offsets or hedges that rose during March of this year?
I would trade VIX the same way you would trade market beta: by holding a constant allocation to the corresponding risk factors and rebalancing regularly. If you don't believe there are risk factors, well, that would indicate market timing or arbitrage, and this ain't arbitrage. :happy
Rebalancing is timing according to you. The fact that you want to wrap the word "beta" around it changes nothing. Beta rebalancing is removing valuation skew and reversion. I agree with the principle.
My real account doesn't trade options or volatility because I haven't been able to uncover any strong evidence indicating this would result in higher CER. All academic evidence points in the opposite direction that you are trading: long volatility is a net loss position and short volatility is a net win position. This is not the result of market inefficiency, but a premium for unhedgeable risk that occurs as a byproduct of delta-hedging options.
LOL. I agree that the idiots who blindly buy call options on some hot stock have no idea what they are doing.
FTR - I don't trade something or not trade something because of blindly adhering to one snippet of "academic evidence". There is plenty of evidence describing low probability high impact events that are unknown but can and do yield fortunes to those who understand the theory. How else do you explain guys like Taleb who have literally made fortunes by using option tranches to benefit in high impact low probability events? The VIX hedge that I use does the same. It's principle-based. You know the same principle that is used by every major company that employs multiple data centers to subvert low probability high impact events.

I fully understand selling volatility. I sell premium on thousands of options nearly every day. But I understand what low probability high impact events (black swans) are and in this case how to benefit from market crashes. Perhaps you might look at the data and understand how to benefit from Black Swans (yeah they are inefficient market events) and understand how to play them. Or do you dispute that since the market is efficient that it won't crash again? Lots of book smart guys saw their a$$eS in 1998 because they thought the efficient market wouldn’t crash. I'm sure you are familiar with their story.
User avatar
Uncorrelated
Posts: 1235
Joined: Sun Oct 13, 2019 3:16 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Uncorrelated »

perfectuncertainty wrote: Tue Dec 01, 2020 8:15 am You advocate for rebalancing. Why rebalance? Because valuation in your portfolio is skewed? Makes sense - but according to you, that would be timing. If you rebalance to align with probability - that's precisely why I don't buy VIX options that will not have value in the event they have little to no probability of having multiples of intrinsic value if the VIX retests its highs.
Rebalancing is not timing. If you have determined that the optimal beta exposure is 0.7 and the market drops, your beta exposure might change to 0.6. There are two actions you can take: you can stick to your strategy and rebalance, or you can make the conscious decision that 0.6 beta is better than 0.7, even though you said the optimal exposure was 0.7 last week. The latter is market timing because it can only be rationalized if you believe market expectations have changed significantly.

The same holds for insurance. Suppose you hold a home and you believe it's smart to purchase fire insurance. If your house burns down and you buy a new house, would you not purchase fire insurance? Of course you would, the arguments why you previously believe fire insurance was optimal still apply. You don't wait until your original fire insurance contract would have expired before purchasing new insurance.

Also for options. Suppose you have determined holding a put with a strike 10% below spot is optimal. The stock market goes up 20%, so you're now holding a put with a strike price 30% below spot. This position is not optimal, so you should rebalance back to the optimal portfolio. If you don't do this, that would be a contradiction of the original claim that 10% below spot is the optimal put.

All of this traces back to generic optimization principles. If your expectations are constant then your asset allocation should also be constant. If your expectations are time varying, then your asset allocation is also time-varying, which is what I define as market timing.

The market is not always efficient - depending on your definition of the market. I know your book says it is, but tell me why massive backwardation is currently occurring in Tesla call options right now? Market efficiency? I'd love to hear your answer to this question.
If the market is not efficient, it should be fairly trivial to prove. Even the biggest supporters of behavioral market models have so far been unable to formulate proofs of market inefficiency. Sure the market isn't 100% efficient, but they're not massively inefficient all the time.

Implied volatility backwardation is a normal phenomena that occurs due to volatility clustering. In normal times the expected volatility (near month to far month) might be 10%, 11%, 11% .... and long traders that roll lose money as future contracts are higher priced. In turbulent times, the volatility (near month to far month) might be 50%, 30%, 20%.... eventually settling down to 11% (the same as normal times). Remember that the implied volatility is the dispersion in outcomes, over long time horizons the dispersion converges to some equilibrium regardless of whether we are currently experiencing elevated volatility or not.
What you're really saying is that you can predict the VIX with greater accuracy than the market.

Would love to see any research indicating that this is possible. If you can predict the VIX better than the market, that would pretty clearly indicate that the markets are inefficient. Where are all the active managers making money on VIX futures? If you are making money, who is on the other side of that trade?

If you perform a call or put strategy (for example: put with strike 10% below current price) and the price moves such that the put no longer has a strike 10% below current price, you should probably rebalance if you want to avoid market timing and anchoring. I suppose you could call that "cashing out", but I call it rebalancing. I explained my reasoning in this topic I opened last week. This might sound contradictory (after all, this way you'll never capitalize on the put...), but is the only strategy consistent with some very fundamental mathematical theories. Unless you believe you can time the markets of course, then the theory goes out of the window.
Nope. Not making any predictions. I do not know when the VIX is going to spike (stated such multiple times). I do know that reversion of the VIX occurs quite quickly and that when I have multiple tens of thousands of options that are deep ITM and the VIX is over 65 I will gladly cash out every time.

Or do you dispute VIX reversion?
I don't dispute VIX reversion. I dispute what you're implying with your trading strategy: that a bunch of complete idiots are going long VIX when everybody knows it's going to reverse. Of course that cannot not true, the market might not be perfectly efficient but it's not widely inefficient either. The only conclusion is that the expected reversion is priced in the futures price.

If you're profiting, you always have to ask who's at the other side of the trade. You are a net volatility buyer. Large algo traders (delta hedgers) are net volatility sellers. Which side do you think is more sophisticated?
LOL. I agree that the idiots who blindly buy call options on some hot stock have no idea what they are doing.
FTR - I don't trade something or not trade something because of blindly adhering to one snippet of "academic evidence". There is plenty of evidence describing low probability high impact events that are unknown but can and do yield fortunes to those who understand the theory. How else do you explain guys like Taleb who have literally made fortunes by using option tranches to benefit in high impact low probability events? The VIX hedge that I use does the same. It's principle-based. You know the same principle that is used by every major company that employs multiple data centers to subvert low probability high impact events.

I fully understand selling volatility. I sell premium on thousands of options nearly every day. But I understand what low probability high impact events (black swans) are and in this case how to benefit from market crashes. Perhaps you might look at the data and understand how to benefit from Black Swans (yeah they are inefficient market events) and understand how to play them. Or do you dispute that since the market is efficient that it won't crash again? Lots of book smart guys saw their a$$eS in 1998 because they thought the efficient market wouldn’t crash. I'm sure you are familiar with their story.
There is plenty of evidence that purchasing tail insurance is a net loss position. For example see Downside Variance Risk Premium or Quantifying the Variance Risk Premium in VIX Options. The interesting part is that you're claiming the exact opposite, but I just can't find any papers supporting that position. Do you have any?

Efficient markets can crash for various reasons. Markets are efficient, not clairvoyant.
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

Uncorrelated wrote: Tue Dec 01, 2020 10:35 am
perfectuncertainty wrote: Tue Dec 01, 2020 8:15 am You advocate for rebalancing. Why rebalance? Because valuation in your portfolio is skewed? Makes sense - but according to you, that would be timing. If you rebalance to align with probability - that's precisely why I don't buy VIX options that will not have value in the event they have little to no probability of having multiples of intrinsic value if the VIX retests its highs.
Rebalancing is not timing. If you have determined that the optimal beta exposure is 0.7 and the market drops, your beta exposure might change to 0.6. There are two actions you can take: you can stick to your strategy and rebalance, or you can make the conscious decision that 0.6 beta is better than 0.7, even though you said the optimal exposure was 0.7 last week. The latter is market timing because it can only be rationalized if you believe market expectations have changed significantly.

The same holds for insurance. Suppose you hold a home and you believe it's smart to purchase fire insurance. If your house burns down and you buy a new house, would you not purchase fire insurance? Of course you would, the arguments why you previously believe fire insurance was optimal still apply. You don't wait until your original fire insurance contract would have expired before purchasing new insurance.

Also for options. Suppose you have determined holding a put with a strike 10% below spot is optimal. The stock market goes up 20%, so you're now holding a put with a strike price 30% below spot. This position is not optimal, so you should rebalance back to the optimal portfolio. If you don't do this, that would be a contradiction of the original claim that 10% below spot is the optimal put.

All of this traces back to generic optimization principles. If your expectations are constant then your asset allocation should also be constant. If your expectations are time varying, then your asset allocation is also time-varying, which is what I define as market timing.

The market is not always efficient - depending on your definition of the market. I know your book says it is, but tell me why massive backwardation is currently occurring in Tesla call options right now? Market efficiency? I'd love to hear your answer to this question.
If the market is not efficient, it should be fairly trivial to prove. Even the biggest supporters of behavioral market models have so far been unable to formulate proofs of market inefficiency. Sure the market isn't 100% efficient, but they're not massively inefficient all the time.

Implied volatility backwardation is a normal phenomena that occurs due to volatility clustering. In normal times the expected volatility (near month to far month) might be 10%, 11%, 11% .... and long traders that roll lose money as future contracts are higher priced. In turbulent times, the volatility (near month to far month) might be 50%, 30%, 20%.... eventually settling down to 11% (the same as normal times). Remember that the implied volatility is the dispersion in outcomes, over long time horizons the dispersion converges to some equilibrium regardless of whether we are currently experiencing elevated volatility or not.
What you're really saying is that you can predict the VIX with greater accuracy than the market.

Would love to see any research indicating that this is possible. If you can predict the VIX better than the market, that would pretty clearly indicate that the markets are inefficient. Where are all the active managers making money on VIX futures? If you are making money, who is on the other side of that trade?

If you perform a call or put strategy (for example: put with strike 10% below current price) and the price moves such that the put no longer has a strike 10% below current price, you should probably rebalance if you want to avoid market timing and anchoring. I suppose you could call that "cashing out", but I call it rebalancing. I explained my reasoning in this topic I opened last week. This might sound contradictory (after all, this way you'll never capitalize on the put...), but is the only strategy consistent with some very fundamental mathematical theories. Unless you believe you can time the markets of course, then the theory goes out of the window.
Nope. Not making any predictions. I do not know when the VIX is going to spike (stated such multiple times). I do know that reversion of the VIX occurs quite quickly and that when I have multiple tens of thousands of options that are deep ITM and the VIX is over 65 I will gladly cash out every time.

Or do you dispute VIX reversion?
I don't dispute VIX reversion. I dispute what you're implying with your trading strategy: that a bunch of complete idiots are going long VIX when everybody knows it's going to reverse. Of course that cannot not true, the market might not be perfectly efficient but it's not widely inefficient either. The only conclusion is that the expected reversion is priced in the futures price.

If you're profiting, you always have to ask who's at the other side of the trade. You are a net volatility buyer. Large algo traders (delta hedgers) are net volatility sellers. Which side do you think is more sophisticated?
LOL. I agree that the idiots who blindly buy call options on some hot stock have no idea what they are doing.
FTR - I don't trade something or not trade something because of blindly adhering to one snippet of "academic evidence". There is plenty of evidence describing low probability high impact events that are unknown but can and do yield fortunes to those who understand the theory. How else do you explain guys like Taleb who have literally made fortunes by using option tranches to benefit in high impact low probability events? The VIX hedge that I use does the same. It's principle-based. You know the same principle that is used by every major company that employs multiple data centers to subvert low probability high impact events.

I fully understand selling volatility. I sell premium on thousands of options nearly every day. But I understand what low probability high impact events (black swans) are and in this case how to benefit from market crashes. Perhaps you might look at the data and understand how to benefit from Black Swans (yeah they are inefficient market events) and understand how to play them. Or do you dispute that since the market is efficient that it won't crash again? Lots of book smart guys saw their a$$eS in 1998 because they thought the efficient market wouldn’t crash. I'm sure you are familiar with their story.
There is plenty of evidence that purchasing tail insurance is a net loss position. For example see Downside Variance Risk Premium or Quantifying the Variance Risk Premium in VIX Options. The interesting part is that you're claiming the exact opposite, but I just can't find any papers supporting that position. Do you have any?

Efficient markets can crash for various reasons. Markets are efficient, not clairvoyant.
Oh please.

I'm not going to put much more effort into this. You choose to ignore anything that doesn't fit your bias.

FTR, I'll state my views on the efficiency of the market:
1. The market is very efficient in the long term.
2. The market is generally efficient in the short term.

As far as timing you failed to answer why backwardation occurs in an efficient market. Specifically, I asked you why Tesla call options are experiencing such backwardation currently (RIGHT NOW the call options in December have an IVx of 109.5% and July 2021 has an IVx of 68.4%). I'm not talking about some theory - this is the reality currently. How does the EMH account for this?

As far as quoting papers, we've all read them, I quoted hundreds in my dissertation. The fact that no one has chosen to study a particular scenario does not invalidate the logic or the results of a scenario. Or are you claiming that the VIX hedge I use is disproven in your paper? My account would prove the inverse to be true. However, you might claim that my sample set is low. Got that. I do know the VIX hedge worked in 2008 (I didn't have it on then, but backtested it) and in 2020 (with the hedge up over 6,900% when I cashed it). Again, the EMH doesn't account for low probability / high impact events, however, myself and many others have benefitted from them and not because of luck, but because of thoughtful preparation and execution. Again, you misquote and obviously don't understand the VIX hedge approach. I'm not re-explaining it for you. All I'll state is that I always have the hedge on and I don't know when the low probability event is going to occur but I'll benefit from the impact because I always have a line in that water.

It's almost comical to hear you state that rebalancing is not timing yet at the same time you should ensure that your exposure is optimal. What is even funnier is that is precisely the logic behind the hedge I use. Always be exposed optimally. You don't know how the market is going to react and how any segment in your balancing is going to perform so you keep rebalancing to ensure exposure so that when reversion occurs you are poised to benefit. Same principle. My VIX hedge allocation is constantly based on the Net Liq in my account (yeah academia doesn't discuss that theory). I'd brand it as using "street smart" "generic optimization principles" :-)
User avatar
Uncorrelated
Posts: 1235
Joined: Sun Oct 13, 2019 3:16 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Uncorrelated »

perfectuncertainty wrote: Tue Dec 01, 2020 11:22 am I'm not going to put much more effort into this. You choose to ignore anything that doesn't fit your bias.
Perhaps it would be helpful if you provided some academic sources that challenge my "biases". It's hard to change a man's mind if you systematically ignore my requests for evidence.

As far as timing you failed to answer why backwardation occurs in an efficient market. Specifically, I asked you why Tesla call options are experiencing such backwardation currently (RIGHT NOW the call options in December have an IVx of 109.5% and July 2021 has an IVx of 68.4%). I'm not talking about some theory - this is the reality currently. How does the EMH account for this?
This is a natural consequence of market uncertainty. If the market is uncertain, it's likely to be uncertain tomorrow, therefore high volatility predicts future high volatility. If the market is certain, it's likely to be certain tomorrow, therefore low volatility predicts future low volatility. But this only works over short term horizons. Over sufficiently long time horizons the best estimate for future volatility is the steady state volatility.

In case you're asking why Tesla's volatility is currently high, I don't know, I don't follow individual stocks. I do know that this is exactly what you would expect from an efficient market operating under uncertain information. It's not anomalous, not indicative of of market inefficiency, and not exploitable.

As far as quoting papers, we've all read them, I quoted hundreds in my dissertation. The fact that no one has chosen to study a particular scenario does not invalidate the logic or the results of a scenario. Or are you claiming that the VIX hedge I use is disproven in your paper? My account would prove the inverse to be true. However, you might claim that my sample set is low. Got that. I do know the VIX hedge worked in 2008 (I didn't have it on then, but backtested it) and in 2020 (with the hedge up over 6,900% when I cashed it). Again, the EMH doesn't account for low probability / high impact events, however, myself and many others have benefitted from them and not because of luck, but because of thoughtful preparation and execution. Again, you misquote and obviously don't understand the VIX hedge approach. I'm not re-explaining it for you. All I'll state is that I always have the hedge on and I don't know when the low probability event is going to occur but I'll benefit from the impact because I always have a line in that water.
My paper doesn't invalidate your strategy, but it does strongly hint in the opposite direction. That's why I'm asking you to provide evidence. Your account is not evidence, neither is a backtest to 2008, that's an anecdote. It's not hard at all to find a strategy that works in 2008 and 2020. Wat is hard is finding a strategy that works well for future market conditions.

The EMH does account for low probability/high impact events. On top of my head, the probability of a 10% drop in a month is around 4%. But the way options are priced suggests that the probability is around 11%. Why is the probability so much larger? Because people that purchase insurance are willing to massively overpay. I massively overpay for fire insurance, too, everybody knows it's a good idea to pay a small premium to avoid a colossal life-altering loss. Fire insurance is a net loss position. A put 10% below current price is a net loss position (see paper I linked). You seem to believe a put 50% below the current price is a net win position. If you have read hundreds of papers, could you cite one that explains how it's possible a put 10% below spot is a losing proposition, but a put 50% below spot is a winner?
ChrisDaPepper
Posts: 1
Joined: Tue Dec 01, 2020 1:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by ChrisDaPepper »

SOXL has been my golden goose across two accounts.

Account 1 - Sold some RYVYX and all TQQQ in this account. It's a self-managed 401k that would not allow LIFO, which sucks because my RYVYX would be a lot better. The RYVYX lots that were sold were bought on 04/16 (almost hit 100%), 04/28 (50%+), and 06/08 (50%+).

SOXL
9/22/2020 Short
96.67%
9/16/2020 Short
78.91%
9/3/2020 Short
69.11%
8/28/2020 Short
69.47%
7/10/2020 Short
107.39%
7/9/2020 Short
110.05%
6/11/2020 Short
155.15%
6/3/2020 Short
148.20%
5/18/2020 Short
211.34%

RYVYX
8/17/2020 Short
15.63%
8/7/2020 Short
15.92%
8/3/2020 Short
20.45%
7/20/2020 Short
25.79%
7/7/2020 Short
28.23%
6/25/2020 Short
42.52%
6/22/2020 Short
43.58%
6/8/2020 Short
47.65%
6/8/2020 Short
50.88%

TECL
9/16/2020 Short
23.97%
5/18/2020 Short
98.52%

Will post Account 2 soon.
ksea
Posts: 1
Joined: Tue Dec 01, 2020 2:08 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by ksea »

perfectuncertainty wrote: Tue Dec 01, 2020 8:15 am I do know that reversion of the VIX occurs quite quickly and that when I have multiple tens of thousands of options that are deep ITM and the VIX is over 65 I will gladly cash out every time.
Naïve question:

If this is the core thesis, then why not
1) sell short VIX options (or something tracking the VIX itself, like VXX) when these conditions occur
2) cover when the reversion happens

Then you wouldn't need to hold the hedge the rest of the time.

Is there some mechanism that makes this difficult?
Last edited by ksea on Tue Dec 01, 2020 2:58 pm, edited 1 time in total.
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

Nope. This isn't school. I'm out - I have money to make.
Semantics
Posts: 187
Joined: Tue Mar 10, 2020 1:42 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Semantics »

Uncorrelated wrote: Tue Dec 01, 2020 10:35 am Efficient markets can crash for various reasons. Markets are efficient, not clairvoyant.
Indeed, but doesn't the lack of clairvoyance i.e. uncertainty imply that hedging strategies can work?

There was a Ben Felix video linked in another thread a few days ago about market bubbles that explained what to me (no finance background) was a very powerful and intuitive insight -- since price as a function of earnings growth is convex, Jensen's inequality implies assets will be more overpriced when there's high uncertainty about earnings growth (in the upward direction from fundamentals), assuming the eventual ground truth earnings growth is somewhere closer to the mean.

Seems to me we could apply the same idea to crashes. During a black swan, the market starts to price in some < 1.0 multiplier on the prior earnings expectation. Since there's often high uncertainty around these events, the mean price drop of companies falls below the price that would be dictated by the mean earnings deflator, due to convexity.

Where VIX comes into it is that it's the weighted average option price, so a measurement of uncertainty. Thus, when the market crashes, the higher VIX goes the more underpriced stocks are likely to be. This too is sort of obvious ("if VIX is high it's time to buy"), but the convexity argument helps me understand why. So if you could track VIX at no cost it would work very well, as PV backtests show even with just monthly rebalancing.

The problem with VIX hedge comes from the implementation details, where options/futures usually are expensive due to usually being in contango (convexity is a factor here as well, I think). What perfectuncertainty's strategy seems to do is minimize that cost by buying only really cheap options, such that a rebalance from VIX calls into equities only occurs when equities are highly underpriced. The cumulative cost of the options is essentially linear with time, but the amount stocks are mispriced by is a convex function of the uncertainty. So if you hedge against rare enough events the mispricing should be significant enough to dominate the cost.

My logic here could be completely flawed, but isn't this basically also what Taleb describes?
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

Semantics wrote: Tue Dec 01, 2020 11:21 pm
Uncorrelated wrote: Tue Dec 01, 2020 10:35 am Efficient markets can crash for various reasons. Markets are efficient, not clairvoyant.
Indeed, but doesn't the lack of clairvoyance i.e. uncertainty imply that hedging strategies can work?

There was a Ben Felix video linked in another thread a few days ago about market bubbles that explained what to me (no finance background) was a very powerful and intuitive insight -- since price as a function of earnings growth is convex, Jensen's inequality implies assets will be more overpriced when there's high uncertainty about earnings growth (in the upward direction from fundamentals), assuming the eventual ground truth earnings growth is somewhere closer to the mean.

Seems to me we could apply the same idea to crashes. During a black swan, the market starts to price in some < 1.0 multiplier on the prior earnings expectation. Since there's often high uncertainty around these events, the mean price drop of companies falls below the price that would be dictated by the mean earnings deflator, due to convexity.

Where VIX comes into it is that it's the weighted average option price, so a measurement of uncertainty. Thus, when the market crashes, the higher VIX goes the more underpriced stocks are likely to be. This too is sort of obvious ("if VIX is high it's time to buy"), but the convexity argument helps me understand why. So if you could track VIX at no cost it would work very well, as PV backtests show even with just monthly rebalancing.

The problem with VIX hedge comes from the implementation details, where options/futures usually are expensive due to usually being in contango (convexity is a factor here as well, I think). What perfectuncertainty's strategy seems to do is minimize that cost by buying only really cheap options, such that a rebalance from VIX calls into equities only occurs when equities are highly underpriced. The cumulative cost of the options is essentially linear with time, but the amount stocks are mispriced by is a convex function of the uncertainty. So if you hedge against rare enough events the mispricing should be significant enough to dominate the cost.

My logic here could be completely flawed, but isn't this basically also what Taleb describes?
The hardest part of holding the hedge is the psychological part of watching the options expire worthless for such long stretches of time. I tell myself it's like an IRA investment (I can't touch it so ignore it). Eventually, it pays out - low probability high impact. Yes, along the lines of what Taleb describes.
langlands
Posts: 758
Joined: Wed Apr 03, 2019 10:05 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by langlands »

perfectuncertainty wrote: Wed Dec 02, 2020 1:31 am
Semantics wrote: Tue Dec 01, 2020 11:21 pm
Uncorrelated wrote: Tue Dec 01, 2020 10:35 am Efficient markets can crash for various reasons. Markets are efficient, not clairvoyant.
Indeed, but doesn't the lack of clairvoyance i.e. uncertainty imply that hedging strategies can work?

There was a Ben Felix video linked in another thread a few days ago about market bubbles that explained what to me (no finance background) was a very powerful and intuitive insight -- since price as a function of earnings growth is convex, Jensen's inequality implies assets will be more overpriced when there's high uncertainty about earnings growth (in the upward direction from fundamentals), assuming the eventual ground truth earnings growth is somewhere closer to the mean.

Seems to me we could apply the same idea to crashes. During a black swan, the market starts to price in some < 1.0 multiplier on the prior earnings expectation. Since there's often high uncertainty around these events, the mean price drop of companies falls below the price that would be dictated by the mean earnings deflator, due to convexity.

Where VIX comes into it is that it's the weighted average option price, so a measurement of uncertainty. Thus, when the market crashes, the higher VIX goes the more underpriced stocks are likely to be. This too is sort of obvious ("if VIX is high it's time to buy"), but the convexity argument helps me understand why. So if you could track VIX at no cost it would work very well, as PV backtests show even with just monthly rebalancing.

The problem with VIX hedge comes from the implementation details, where options/futures usually are expensive due to usually being in contango (convexity is a factor here as well, I think). What perfectuncertainty's strategy seems to do is minimize that cost by buying only really cheap options, such that a rebalance from VIX calls into equities only occurs when equities are highly underpriced. The cumulative cost of the options is essentially linear with time, but the amount stocks are mispriced by is a convex function of the uncertainty. So if you hedge against rare enough events the mispricing should be significant enough to dominate the cost.

My logic here could be completely flawed, but isn't this basically also what Taleb describes?
The hardest part of holding the hedge is the psychological part of watching the options expire worthless for such long stretches of time. I tell myself it's like an IRA investment (I can't touch it so ignore it). Eventually, it pays out - low probability high impact. Yes, along the lines of what Taleb describes.
I don't think it's psychologically difficult at all. I've held SPY puts of varying strikes and maturities since March of this year. The ones I sold in March were obviously up massively but all the others I've held since then have expired worthless. As long as I know what all the pieces in the puzzle (my portfolio) are doing, everything is psychologically easy.

What's difficult is coming up with the right hedging strategy in the first place and making sure it covers all contingencies so that you're essentially never surprised by any outcome.
User avatar
Uncorrelated
Posts: 1235
Joined: Sun Oct 13, 2019 3:16 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Uncorrelated »

Semantics wrote: Tue Dec 01, 2020 11:21 pm
Uncorrelated wrote: Tue Dec 01, 2020 10:35 am Efficient markets can crash for various reasons. Markets are efficient, not clairvoyant.
Indeed, but doesn't the lack of clairvoyance i.e. uncertainty imply that hedging strategies can work?

There was a Ben Felix video linked in another thread a few days ago about market bubbles that explained what to me (no finance background) was a very powerful and intuitive insight -- since price as a function of earnings growth is convex, Jensen's inequality implies assets will be more overpriced when there's high uncertainty about earnings growth (in the upward direction from fundamentals), assuming the eventual ground truth earnings growth is somewhere closer to the mean.

Seems to me we could apply the same idea to crashes. During a black swan, the market starts to price in some < 1.0 multiplier on the prior earnings expectation. Since there's often high uncertainty around these events, the mean price drop of companies falls below the price that would be dictated by the mean earnings deflator, due to convexity.

Where VIX comes into it is that it's the weighted average option price, so a measurement of uncertainty. Thus, when the market crashes, the higher VIX goes the more underpriced stocks are likely to be. This too is sort of obvious ("if VIX is high it's time to buy"), but the convexity argument helps me understand why. So if you could track VIX at no cost it would work very well, as PV backtests show even with just monthly rebalancing.

The problem with VIX hedge comes from the implementation details, where options/futures usually are expensive due to usually being in contango (convexity is a factor here as well, I think). What perfectuncertainty's strategy seems to do is minimize that cost by buying only really cheap options, such that a rebalance from VIX calls into equities only occurs when equities are highly underpriced. The cumulative cost of the options is essentially linear with time, but the amount stocks are mispriced by is a convex function of the uncertainty. So if you hedge against rare enough events the mispricing should be significant enough to dominate the cost.

My logic here could be completely flawed, but isn't this basically also what Taleb describes?
To be clear: I never said hedging strategies can't work. I said hedging strategies reduce the expected return by a significant amount. This might result in higher risk adjusted returns, but it might not.

VIX does not measure uncertainty. It measures uncertainty plus the variance risk premium. The variance risk premium varies over time (also perfectly possible in efficient markets). The VIX ends up being a pretty poor measure of uncertainty, it took me a few days with regression models to outperform VIX as a measure of expected volatility quite substantially.

I don't fully understand Ben Felix' argument, but my impression is that if there was a rational basis behind the bubbles, it should also be rational for you to go along with that bubble. You can make money in two ways: have different risk preferences from the market and bet on risk factors, or bet against irrational pricing. I don't see the irrationality here and I also don't see how it's obvious your risk preferences are different from the rest of the market.
Jensen's inequality implies assets will be more overpriced when there's high uncertainty about earnings growth
Thus, when the market crashes, the higher VIX goes the more underpriced stocks are likely to be.
It looks like you contradicted yourself.
perfectuncertainty
Posts: 195
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

langlands wrote: Wed Dec 02, 2020 2:08 am
perfectuncertainty wrote: Wed Dec 02, 2020 1:31 am The hardest part of holding the hedge is the psychological part of watching the options expire worthless for such long stretches of time. I tell myself it's like an IRA investment (I can't touch it so ignore it). Eventually, it pays out - low probability high impact. Yes, along the lines of what Taleb describes.
I don't think it's psychologically difficult at all. I've held SPY puts of varying strikes and maturities since March of this year. The ones I sold in March were obviously up massively but all the others I've held since then have expired worthless. As long as I know what all the pieces in the puzzle (my portfolio) are doing, everything is psychologically easy.

What's difficult is coming up with the right hedging strategy in the first place and making sure it covers all contingencies so that you're essentially never surprised by any outcome.
Great way to look at it.
Walkure
Posts: 307
Joined: Tue Apr 11, 2017 9:59 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Walkure »

As an aside for those of us holding HFEA in taxable, I noticed that both UPRO and TMF have suspended their quarterly dividend policy. The last distribution was for 1Q 2020. Granted they were only yielding .4% previously, so it never mattered much anyway, but good to know.
Thereum
Posts: 62
Joined: Sun Jun 14, 2020 9:05 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Thereum »

I did some more research.

Volatility decay seems to cost about 5% a year with leveraged long ETFs. With leveraged short ETFs, it seems to cost around 10% a year.

Long term options tend to cost much less than 5% a year, so you are better off achieving leverage that way. You can also get margin for under 2% a year, which is another good way to achieve leverage.

Of course, the holy grail is shorting inverse leveraged ETFs, but this requires a bit of active management to limit risk.

Holding leveraged ETFs over the long term is a terrible strategy. And those of you are planning to hold leveraged long term treasuries over the next decade are in for a very rude awakening. You are going to have nothing but headwinds.
Last edited by Thereum on Wed Dec 02, 2020 5:47 pm, edited 1 time in total.
langlands
Posts: 758
Joined: Wed Apr 03, 2019 10:05 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by langlands »

Thereum wrote: Wed Dec 02, 2020 4:22 pm I did some more research.

Volatility decay seems to cost about 5% a year with leveraged long ETFs. With leveraged short ETFs, it seems to cost around 10% a year.

Long term options tend to cost much less than 5% a year, so you are better off achieving leverage that way. You can also get margin for under 2% a year, which is another good way to achieve leverage.

Of course, the holy grail is shorting inverse leveraged ETFs, but this requires a bit of active management to limit risk.

Holding leveraged ETFs over the long term is a terrible strategy. And those of you are planning to hold leveraged long term treasuries over the next decade: May god have mercy on your souls.
How can you make a categorical statement of "5% a year" when you don't specify the leverage amount (2x or 3x) and you don't specify the underlying index (TQQQ, UPRO, JNUG)? In any case, the formula for volatility decay is well-known and explained well here: https://www.bogleheads.org/wiki/Variance_drain. It's just a function of the leverage amount and the volatility of the underlying. And it's ultimately a mathematical fact about the difference between arithmetic (expected value) and geometric (median) returns.

Trying to "arbitrage" the median outcome against the expected outcome is picking up pennies in front of a steamroller, and in the long run you'll be crushed. What you are attempting to do is exactly akin to the martingale strategy in gambling in which you double your bet (increase your leverage) every time you lose. If you have a deep enough bankroll, you'll win 99% of the time. But the one time you lose, you'll lose everything. If you notice this and try to hedge...well guess what that'll cost you. How much does that cost you? The market-efficient price. Volatility decay isn't a "free lunch" to be picked up.
Thereum
Posts: 62
Joined: Sun Jun 14, 2020 9:05 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Thereum »

langlands wrote: Wed Dec 02, 2020 5:17 pm
How can you make a categorical statement of "5% a year" when you don't specify the leverage amount (2x or 3x) and you don't specify the underlying index (TQQQ, UPRO, JNUG)? In any case, the formula for volatility decay is well-known and explained well here: https://www.bogleheads.org/wiki/Variance_drain. It's just a function of the leverage amount and the volatility of the underlying. And it's ultimately a mathematical fact about the difference between arithmetic (expected value) and geometric (median) returns.

Trying to "arbitrage" the median outcome against the expected outcome is picking up pennies in front of a steamroller, and in the long run you'll be crushed. What you are attempting to do is exactly akin to the martingale strategy in gambling in which you double your bet (increase your leverage) every time you lose. If you have a deep enough bankroll, you'll win 99% of the time. But the one time you lose, you'll lose everything. If you notice this and try to hedge...well guess what that'll cost you. How much does that cost you? The market-efficient price. Volatility decay isn't a "free lunch" to be picked up.
I simply short the leveraged ETF and go long on the underlying index (e.g., short TQQQ, long QQQ) and adjust the weights so that the beta equals zero.

I am not recommending arbitrage, although you certainly could do it. What I am advising is to not hold leveraged ETFs over long periods of time (years or more). Leveraged ETFs are such terrible investments that they are best used for bearish bets. I am a huge fan of leverage, however, and the best way to do that is with options (which are cheaper than leveraged ETFs).
Post Reply