HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Tue Jun 11, 2019 4:23 pm

gw wrote:
Tue Jun 11, 2019 3:41 pm
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 10:32 am
coingaroo wrote:
Tue Jun 11, 2019 10:23 am
gw wrote:
Sun Mar 10, 2019 10:11 am
You're missing or evading the basic point, which is that effectively all you've done is to identify that bonds were a great investment over the last thirty years.
...
Thank you for one of the most insightful posts in this thread. Of course you will get insane performance by picking the two best performing asset classes and specifications, that maintain some level of diversification and then leverage like crazy.

Gw, what do you think about the idea of equal weight contribution (risk parity) in general? ERC of stocks, treasuries, bonds, commodities, property; leveraged to a target volatility.
GW’s objection raises the question: what is it reasonable to expect ex-ante? I submit the following:

1. That the equity premium will persist.
2. That the term premium will persist.
3. That flight to safety will persist.

These are risk-based premia that comprise the fundamental drivers of the financial markets. They don’t need to be “discounted”. By relying on them I am certainly not making a “big bet”. Every BH is relying on them with the 3-fund portfolio.

I make no prediction about interest rates; we have seen in the recon past how that has panned out.
Suppose that, over the next few years, the 10-year bond rate were to rise from 2.5% to 5% - essentially just returning to 2000-2009 levels. That seems within the realm of possibilities.

The last time rates doubled was the period 1977-1981. This strategy lost something like 65% over that 4-year period, vs. a 150% gain for the S&P 500.

The previous time rates doubled was the period 1961-1975. This strategy lost something like 50% over that 14-year period, vs. a 200% gain for the S&P 500.

(I'm grabbing those dates and returns by eye from the charts below. YMMV.)

https://fred.stlouisfed.org/series/GS10

https://i.imgur.com/g3fRGNS.png (viewtopic.php?f=10&t=272007&start=1050#p4426381)

Hedgefundie's argument is that the market conditions of the 60s-70s are a thing of the past. Maybe so. Maybe not. I'm just saying he should recognize the extent to which this strategy's historical performance has been juiced by 40 years of steadily falling interest rates.

My general view of risk parity strategies is similar. One has to worry that it's just an elaborate excuse to justify holding more bonds, discovered after a lengthy period over which bonds were an unexpectedly fabulous investment.

Backtests are cheap, and it's easy to fool yourself.
What is the ex-ante reason we should accept sharply rising rates as a serious risk to my strategy? I agree we should base our decisions on ex-ante reasoning, not just backtesting. I've given you my ex-ante reasoning, give me yours.

To the extent that we can form an ex-ante position on interest rates, it's probably that rates will continue to decline or stagnate, given demographic and technology trends.

gw
Posts: 756
Joined: Thu May 28, 2009 10:12 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by gw » Tue Jun 11, 2019 5:05 pm

HEDGEFUNDIE wrote:
Tue Jun 11, 2019 4:23 pm
What is the ex-ante reason we should accept sharply rising rates as a serious risk to my strategy? I agree we should base our decisions on ex-ante reasoning, not just backtesting. I've given you my reasoning, give me yours.

To the extent that we can form an ex-ante position on interest rates, it's probably that rates will continue to decline or stagnate, given demographic and technology trends.
Are you disputing that rising rates are bad for the strategy? The only historical examples were pretty horrible.

It seems unlikely to me that rates will continue to decline in the future as they have in the past (e.g. 10-year from 15% in the 80s to 2.5% now). Opinions may vary, but to me, 2.5% rates vs. a 2% inflation target seems pretty close to a floor.

Anyway, I'm not very interested in trying to forecast rates. If either of us actually knew anything about future rates, we could go make some real money.

The bigger point is that you haven't made any effort to discount the historical outperformance. Sure, you historically would have made a lot of money in levered long bonds over decades of steadily falling rates, and those returns are baked into your (over?)-appealing backtest results. But what are your expected returns going forward, if say, rates stagnate as you suggested? What if they just return to the levels of ten years ago? Is that so implausible? What is your best-case scenario for further long term rate declines, and how does it compare to the alternatives? It just feels like you're not seriously considering these issues.

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Tue Jun 11, 2019 5:17 pm

gw wrote:
Tue Jun 11, 2019 5:05 pm
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 4:23 pm
What is the ex-ante reason we should accept sharply rising rates as a serious risk to my strategy? I agree we should base our decisions on ex-ante reasoning, not just backtesting. I've given you my reasoning, give me yours.

To the extent that we can form an ex-ante position on interest rates, it's probably that rates will continue to decline or stagnate, given demographic and technology trends.
Are you disputing that rising rates are bad for the strategy? The only historical examples were pretty horrible.

It seems unlikely to me that rates will continue to decline in the future as they have in the past (e.g. 10-year from 15% in the 80s to 2.5% now). Opinions may vary, but to me, 2.5% rates vs. a 2% inflation target seems pretty close to a floor.

Anyway, I'm not very interested in trying to forecast rates. If either of us actually knew anything about future rates, we could go make some real money.

The bigger point is that you haven't made any effort to discount the historical outperformance. Sure, you historically would have made a lot of money in levered long bonds over decades of steadily falling rates, and those returns are baked into your (over?)-appealing backtest results. But what are your expected returns going forward, if say, rates stagnate as you suggested? What if they just return to the levels of ten years ago? Is that so implausible? What is your best-case scenario for further long term rate declines, and how does it compare to the alternatives? It just feels like you're not seriously considering these issues.
Rising interest rates are good for the bond portion of the strategy in the long run and bad for the bond portion of the strategy in the short run. How it affects equities is anyone’s guess. With stagnant interest rates the bonds will do just fine, with the yield to maturity staying constant and levered up.

But as you say, predicting rates is an uninteresting exercise. I’m much more interested in the underlying risk premia that make the strategy tick.

gw
Posts: 756
Joined: Thu May 28, 2009 10:12 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by gw » Tue Jun 11, 2019 6:42 pm

HEDGEFUNDIE wrote:
Tue Jun 11, 2019 5:17 pm
Rising interest rates are good for the bond portion of the strategy in the long run and bad for the bond portion of the strategy in the short run.
... where I guess "short run" includes steadily losing half your portfolio over a 14-year period while stocks go up, as in the historical examples above.

I just don't think you've really thought this through, and you've got a number of people following you with their retirements at stake.

MotoTrojan
Posts: 4528
Joined: Wed Feb 01, 2017 8:39 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by MotoTrojan » Tue Jun 11, 2019 7:06 pm

gw wrote:
Tue Jun 11, 2019 6:42 pm
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 5:17 pm
Rising interest rates are good for the bond portion of the strategy in the long run and bad for the bond portion of the strategy in the short run.
... where I guess "short run" includes steadily losing half your portfolio over a 14-year period while stocks go up, as in the historical examples above.

I just don't think you've really thought this through, and you've got a number of people following you with their retirements at stake.
All I have to say is I really enjoy this back and forth of opinions and hole poking, so please do keep it up!

One thing I find interesting is the yield curve spread, as that is a big driver of how much the rising rates help you in the long run. Curious how that compared in the past to more modern monetary periods.

User avatar
305pelusa
Posts: 227
Joined: Fri Nov 16, 2018 10:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by 305pelusa » Tue Jun 11, 2019 7:36 pm

I'm with gw on this one. LIBOR rate is at 2.41%. Current long term treasury yields are at 2.875%. That's a term premium of 0.465%. The expense ratio is ~1%. So investing 10k in TMF should net (if rates stay constant):

(Return of the unlevered part) + (Return of levered part) - (Expense) = Returns
10000*(1.0287) + 20000*(.00465) - 100 =10280

(At least I think that's right... maybe it's incorrect though)
As in 2.8% return annually, lower than if you had just purchased the LT bonds themselves. Not to mention the volatility is three times higher.

Other way to think about it: Borrowing at ST rates to invest at LT rates should enhance the term premium but the expense ratio seems so high that it gets nulled. I think you'd need to leverage far more before the term premium shows itself above the expense ratio.

I'm with gw on this one: If TMF has shown excellent results, I think it's because of the consistent lowering of interest rates during the backtest which add capital gains to the above

I must be missing something obvious about this based on the quantity of posters putting thousands of dollars on the line.

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Tue Jun 11, 2019 7:38 pm

gw wrote:
Tue Jun 11, 2019 6:42 pm
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 5:17 pm
Rising interest rates are good for the bond portion of the strategy in the long run and bad for the bond portion of the strategy in the short run.
... where I guess "short run" includes steadily losing half your portfolio over a 14-year period while stocks go up, as in the historical examples above.

I just don't think you've really thought this through, and you've got a number of people following you with their retirements at stake.
From the beginning, I have framed this strategy as an alternative to a 100% equity portfolio. That is the benchmark by which we should judge this strategy:

1. In how many states of the world would a 100% equity strategy be destroyed?
2. In how many states of the world would my strategy be destroyed?

I think it’s clear that my strategy is robust to more states of the world than a 100% equity strategy.

But you seem to be saying that this strategy should be bulletproof under all possible states of the world. What other investment strategy is held to that standard?

As far as people making this a core part of their retirement portfolios, please point out any of my posts where I have encouraged or condoned such behavior, and I will correct it immediately.

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Tue Jun 11, 2019 7:40 pm

305pelusa wrote:
Tue Jun 11, 2019 7:36 pm
I'm with gw on this one. LIBOR rate is at 2.41%. Current long term treasury yields are at 2.875%. That's a term premium of 0.465%. The expense ratio is ~1%. So investing 10k in TMF should net (if rates stay constant):

(Return of the unlevered part) + (Return of levered part) - (Expense) = Returns
10000*(1.0287) + 20000*(.00465) - 100 =10280

(At least I think that's right... maybe it's incorrect though)
As in 2.8% return annually, lower than if you had just purchased the LT bonds themselves. Not to mention the volatility is three times higher.

Other way to think about it: Borrowing at ST rates to invest at LT rates should enhance the term premium but the expense ratio seems so high that it gets nulled. I think you'd need to leverage far more before the term premium shows itself above the expense ratio.

I'm with gw on this one: If TMF has shown excellent results, I think it's because of the consistent lowering of interest rates during the backtest which add capital gains to the above

I must be missing something obvious about this based on the quantity of posters putting thousands of dollars on the line.
Here is something you are missing - the financing spread on TMF is actually negative, i.e. the financing rate is below LIBOR. You can confirm this via the latest annual reports.

When interest rates are expected to rise, it makes sense that TMF can get paid to take on that risk via swaps.

User avatar
ThereAreNoGurus
Posts: 181
Joined: Fri Jan 24, 2014 11:41 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by ThereAreNoGurus » Tue Jun 11, 2019 7:47 pm

As far as people making this a core part of their retirement portfolios, please point out any of my posts where I have encouraged or condoned such behavior, and I will correct it immediately.
Exactly.

I'm enjoying the back and forth concerning the merits and soundness of the strategy, but the following is not your fault at all and should not discourage you or anybody else from posting (especially since you have posted sufficient caveats) :
[Y]ou've got a number of people following you with their retirements at stake.

User avatar
305pelusa
Posts: 227
Joined: Fri Nov 16, 2018 10:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by 305pelusa » Tue Jun 11, 2019 8:43 pm

HEDGEFUNDIE wrote:
Tue Jun 11, 2019 7:40 pm
305pelusa wrote:
Tue Jun 11, 2019 7:36 pm
I'm with gw on this one. LIBOR rate is at 2.41%. Current long term treasury yields are at 2.875%. That's a term premium of 0.465%. The expense ratio is ~1%. So investing 10k in TMF should net (if rates stay constant):

(Return of the unlevered part) + (Return of levered part) - (Expense) = Returns
10000*(1.0287) + 20000*(.00465) - 100 =10280

(At least I think that's right... maybe it's incorrect though)
As in 2.8% return annually, lower than if you had just purchased the LT bonds themselves. Not to mention the volatility is three times higher.

Other way to think about it: Borrowing at ST rates to invest at LT rates should enhance the term premium but the expense ratio seems so high that it gets nulled. I think you'd need to leverage far more before the term premium shows itself above the expense ratio.

I'm with gw on this one: If TMF has shown excellent results, I think it's because of the consistent lowering of interest rates during the backtest which add capital gains to the above

I must be missing something obvious about this based on the quantity of posters putting thousands of dollars on the line.
Here is something you are missing - the financing spread on TMF is actually negative, i.e. the financing rate is below LIBOR. You can confirm this via the latest annual reports.

When interest rates are expected to rise, it makes sense that TMF can get paid to take on that risk via swaps.
Earlier you mentioned it was ~1.83% when the LIBOR was at 2.33%. The ER is actually more like 1.09%. So if we redo the above

10000*(1.0287) + 20000*(0.01045) - 109 = 10387

That's a return of 3.87%. So you triple the volatility and get an additional 35% over an LT bond fund (like EDV).

Is that right then? That's not a deal I would take. Am I missing something else?

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Tue Jun 11, 2019 8:57 pm

305pelusa wrote:
Tue Jun 11, 2019 8:43 pm
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 7:40 pm
305pelusa wrote:
Tue Jun 11, 2019 7:36 pm
I'm with gw on this one. LIBOR rate is at 2.41%. Current long term treasury yields are at 2.875%. That's a term premium of 0.465%. The expense ratio is ~1%. So investing 10k in TMF should net (if rates stay constant):

(Return of the unlevered part) + (Return of levered part) - (Expense) = Returns
10000*(1.0287) + 20000*(.00465) - 100 =10280

(At least I think that's right... maybe it's incorrect though)
As in 2.8% return annually, lower than if you had just purchased the LT bonds themselves. Not to mention the volatility is three times higher.

Other way to think about it: Borrowing at ST rates to invest at LT rates should enhance the term premium but the expense ratio seems so high that it gets nulled. I think you'd need to leverage far more before the term premium shows itself above the expense ratio.

I'm with gw on this one: If TMF has shown excellent results, I think it's because of the consistent lowering of interest rates during the backtest which add capital gains to the above

I must be missing something obvious about this based on the quantity of posters putting thousands of dollars on the line.
Here is something you are missing - the financing spread on TMF is actually negative, i.e. the financing rate is below LIBOR. You can confirm this via the latest annual reports.

When interest rates are expected to rise, it makes sense that TMF can get paid to take on that risk via swaps.
Earlier you mentioned it was ~1.83% when the LIBOR was at 2.33%. The ER is actually more like 1.09%. So if we redo the above

10000*(1.0287) + 20000*(0.01045) - 109 = 10387

That's a return of 3.87%. So you triple the volatility and get an additional 35% over an LT bond fund (like EDV).

Is that right then? That's not a deal I would take. Am I missing something else?
It's actually only doubling the volatility of EDV, but anyway...

You're missing the purpose of TMF in the strategy as a whole. The volatility is a positive: combined with the negative correlation, the volatility is meant to serve as insurance for an equity crash. TMF delivers slightly more return than EDV during normal times, and a significantly higher rebalancing bonus when equities crash.

EDIT: as far as returns go, if you look at annual returns TMF does return about 2x EDV:

https://www.portfoliovisualizer.com/fun ... hlyReturns

User avatar
coingaroo
Posts: 14
Joined: Fri Apr 26, 2019 11:31 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by coingaroo » Tue Jun 11, 2019 9:23 pm

HEDGEFUNDIE wrote:
Tue Jun 11, 2019 5:17 pm
gw wrote:
Tue Jun 11, 2019 5:05 pm
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 4:23 pm
What is the ex-ante reason we should accept sharply rising rates as a serious risk to my strategy? I agree we should base our decisions on ex-ante reasoning, not just backtesting. I've given you my reasoning, give me yours.

To the extent that we can form an ex-ante position on interest rates, it's probably that rates will continue to decline or stagnate, given demographic and technology trends.
Are you disputing that rising rates are bad for the strategy? The only historical examples were pretty horrible.

It seems unlikely to me that rates will continue to decline in the future as they have in the past (e.g. 10-year from 15% in the 80s to 2.5% now). Opinions may vary, but to me, 2.5% rates vs. a 2% inflation target seems pretty close to a floor.

Anyway, I'm not very interested in trying to forecast rates. If either of us actually knew anything about future rates, we could go make some real money.

The bigger point is that you haven't made any effort to discount the historical outperformance. Sure, you historically would have made a lot of money in levered long bonds over decades of steadily falling rates, and those returns are baked into your (over?)-appealing backtest results. But what are your expected returns going forward, if say, rates stagnate as you suggested? What if they just return to the levels of ten years ago? Is that so implausible? What is your best-case scenario for further long term rate declines, and how does it compare to the alternatives? It just feels like you're not seriously considering these issues.
Rising interest rates are good for the bond portion of the strategy in the long run and bad for the bond portion of the strategy in the short run. How it affects equities is anyone’s guess. With stagnant interest rates the bonds will do just fine, with the yield to maturity staying constant and levered up.

But as you say, predicting rates is an uninteresting exercise. I’m much more interested in the underlying risk premia that make the strategy tick.
Whatever risk premia you get from equity and term will be neutered and turned negative from the expense ratio, swap premium, and volatility drag.

Your strategy is analogous to saying a 40/60 stock and long term treasury allocation have fared well in the past with low drawdowns, meaning you can 3x it, giving you roughly 2x the returns with 3x the risk and drawdowns.

That’s not a good trade, and 60% long term treasuries will certainly do poorly in some regimes.

If you truly want to make a leveraged,regime independent portfolio, you should either look into adding trend following futures to the mix.

Managed futures (trend following) exhibit a ‘smile’ and generate positive returns during stock market crashes, and positive returns over the long run, independent of interest rates movements.

A 3x leveraged, 35% stock, 35% long term treasuries, 20% managed futures allocation would fare substantially better historically — and expected to win in a rising interest rate environment too. You can access this as a retail investor via IB portfolio margin.

User avatar
305pelusa
Posts: 227
Joined: Fri Nov 16, 2018 10:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by 305pelusa » Tue Jun 11, 2019 9:29 pm

HEDGEFUNDIE wrote:
Tue Jun 11, 2019 8:57 pm

It's actually only doubling the volatility of EDV, but anyway...
Could you explain how that's the case? A 3x leveraged ETF will display a 3x change in daily price over the regular fund. If volatility is measured as standard deviation (which I believe is the case), then multiplying a random variable's values by 3 will produce a random variable with 3 times the standard deviation.

History and backtests could always show something different. But to first order, this is approximately what should happen no?
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 8:57 pm

You're missing the purpose of TMF in the strategy as a whole. The volatility is a positive: combined with the negative correlation, the volatility is meant to serve as insurance for an equity crash. TMF delivers slightly more return than EDV during normal times, and a significantly higher rebalancing bonus when equities crash.
Ok so it relies on that negative correlation right? If that's not the case, then you get the result (roughly) I calculated above yes?

I guess that's the part I don't like.
Equity premium? I'd bet money on it.
Term premium? I'd bet money on it.
That a specific correlation stays sufficiently negative between two assets in the future? I would most definitely not bet money on that one.
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 8:57 pm
EDIT: as far as returns go, if you look at annual returns TMF does return about 2x EDV:

https://www.portfoliovisualizer.com/fun ... hlyReturns
Before I reference anything from that graph: the above shows a period of falling interest rates. Since we don't know if they will go up or down, I think the fair comparison is to assume they stay constant. The ER seems soo high, while the term premium so low, that it doesn't seem to me like a good idea.

This is what gw and I keep trying to tell you. We keep saying "but we're not sure this would do well at all if rates stayed constant or went up. It only looks good because rates have been going down".

To which you respond with back test examples of periods with rates going down? Come on man throw us a bone here.

Like gw asked you:
gw wrote:
Tue Jun 11, 2019 5:05 pm
But what are your expected returns going forward, if say, rates stagnate as you suggested? What if they just return to the levels of ten years ago? Is that so implausible? What is your best-case scenario for further long term rate declines, and how does it compare to the alternatives? It just feels like you're not seriously considering these issues.
Is this something that has been tackled on this thread? It's a long way so just point me to the page/poster if so (if not too much trouble!).

gw
Posts: 756
Joined: Thu May 28, 2009 10:12 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by gw » Tue Jun 11, 2019 9:34 pm

HEDGEFUNDIE wrote:
Tue Jun 11, 2019 7:38 pm
From the beginning, I have framed this strategy as an alternative to a 100% equity portfolio. That is the benchmark by which we should judge this strategy:
In your longest available backtest, it has twice the std dev of 100% equities, a 1.5x worse max drawdown (75%), and essentially the same returns (10% vs. 11%), so a pretty big fail.

But worse than that, the strategy essentially *only* made money during the post-1981 period of steadily falling rates - it steadily *lost* money throughout the previous period, when rates were rising.

https://i.imgur.com/9D8QjKf.png

Your response is to simply discard the (long!) period where your strategy underperformed, while keeping all the returns from the period where it outperformed, under an argument that rising rates are a thing of the past. You may be right, but I say it's a big bet.

Again, I would point out that your strategy is roughly equivalent to (100% equities) + 2*(LTT-CASH). It's inherently riskier than 100% equities under any reasonable expectations, has a huge exposure to rising interest rates, and it seems pretty obvious to me why it happened to perform better in the past than one could reasonably expect going forward.

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Tue Jun 11, 2019 9:38 pm

coingaroo wrote:
Tue Jun 11, 2019 9:23 pm

Your strategy is analogous to saying a 40/60 stock and long term treasury allocation have fared well in the past with low drawdowns, meaning you can 3x it, giving you roughly 2x the returns with 3x the risk and drawdowns.

That’s not a good trade, and 60% long term treasuries will certainly do poorly in some regimes.
Take a look at the original backtest, over 30 years it delivered 7x the returns, 2x the risk, and 1x the max drawdown. Find me a better trade than that.
If you truly want to make a leveraged,regime independent portfolio, you should either look into adding trend following futures to the mix.

Managed futures (trend following) exhibit a ‘smile’ and generate positive returns during stock market crashes, and positive returns over the long run, independent of interest rates movements.

A 3x leveraged, 35% stock, 35% long term treasuries, 20% managed futures allocation would fare substantially better historically — and expected to win in a rising interest rate environment too. You can access this as a retail investor via IB portfolio margin.
Interesting idea, feel free to start your own thread. My suspicion, however, is that you’ll run into the same buzzsaw of criticism that I have over the course of the last 2000 posts. “Managed futures are opaque derivatives that deserve no place in a BH portfolio” so on and so forth...

The Broom
Posts: 3
Joined: Tue Jun 11, 2019 8:39 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by The Broom » Tue Jun 11, 2019 10:01 pm

HEDGEFUNDIE, I can't thank you enough for the incredible and thoughtful responses that you've provided in this thread. I've been stalking this site for years and just recently created an account. I've listened to many of Bogle's insights, and have been both elated and overwhelmed by the vast knowledge of the members within this community. It was because of this community that I set up my 3-fund portfolio within my 401k.

What I am failing to understand from the critical comments on HEDGEFUNDIE's adventure is the following:

1. Why are you so negative on this strategy? Is it because you want something with zero risk? Is it because you are looking for something with very limited drawdowns?

I recently re-read through some of the 2008(9) commentary on the forums and saw how members were feeling at the time. It seemed like folks posting at the time were in a state of despair, even though they religiously follow the principles outlined by this investment community. This dispair came from fear -- fear of the unknown, losing everything, fear that the market would never rise again, or worse.

2. Is anyone truly going "all in" on this?

Disclaimer: I'm not!

For me personally, I have discretionary accounts where I make alternative investments (e.g. Bitcoin, TQQQ/TECL longs, sector-specific stocks, and now this strategy). Although many of my alternative investments have been wins, I have made some losses (still holding the bag on a large BTC stake, so I can relate to the HODL mentality that HEDGEFUNDIE mentioned). In terms of percentages, the wins have outnumbered my losses, and I am young enough where I can keep generating new capital for at least the next 3-4 decades to continue to make alternative investments before I get to the point where I'll stop generating capital. I've never seen anyone get to the level that they want to be at without taking a risk. This strategy isn't bulletproof. It's risky. I don't think any of us would disagree with the inherent risk of this.

3. Regarding risk - does anyone truly believe that this would go to zero?

If you do believe that it will hit zero, can you explain why? If this becomes worth nothing, what happens to my 3-fund portfolio? In terms of losses, wouldn't my aggregate losses be substantial in a 3-fund portfolio by the time this hit nothing? Is there a scenario where this goes to zero and the 3-fund portfolio doesn't lose a substantial amount?

--

I ask these questions because I'm trying to understand the negative sentiment that I'm getting over the past few pages here from some of the posters.

Thanks in advance

User avatar
305pelusa
Posts: 227
Joined: Fri Nov 16, 2018 10:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by 305pelusa » Tue Jun 11, 2019 10:35 pm

The Broom wrote:
Tue Jun 11, 2019 10:01 pm
HEDGEFUNDIE, I can't thank you enough for the incredible and thoughtful responses that you've provided in this thread. I've been stalking this site for years and just recently created an account. I've listened to many of Bogle's insights, and have been both elated and overwhelmed by the vast knowledge of the members within this community. It was because of this community that I set up my 3-fund portfolio within my 401k.

What I am failing to understand from the critical comments on HEDGEFUNDIE's adventure is the following:

1. Why are you so negative on this strategy? Is it because you want something with zero risk? Is it because you are looking for something with very limited drawdowns?
For me, it really just comes down to the high expense ratio involved with the strategy. Look, the strategy simply shorts cash and goes long on LT bonds and equities. Intuition tells me this is more risk and also more return. All good so far. For a 1% annual fee of course. That's the part I dislike. To pay those kinds of costs, I'd want something truly remarkable, like one of those AQR funds that invest in alternative lending, insurance, etc that are basically asset classes onto themselves.

The large decrease of interest rates in the past 3 decades seem to mask these costs and make them look worthwhile. But scroll up to my post for an expected return if rates stay the same. It doesn't look good to me at all. Let alone if rates were to go up! My prediction is that these costs will seriously cut into the returns for the decade to come.
The Broom wrote:
Tue Jun 11, 2019 10:01 pm

2. Is anyone truly going "all in" on this?
Not going all in is prudent of course. Idk about you though but I work darn hard for my money. I wouldn't put any significant money in a strategy with a 1% fee. Certainly not any from my precious and limited Roth space.
The Broom wrote:
Tue Jun 11, 2019 10:01 pm

3. Regarding risk - does anyone truly believe that this would go to zero?
I do not think so and it is not my concern at all.

Hopefully the above clarify my concerns.

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Tue Jun 11, 2019 11:05 pm

305pelusa wrote:
Tue Jun 11, 2019 9:29 pm
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 8:57 pm

It's actually only doubling the volatility of EDV, but anyway...
Could you explain how that's the case? A 3x leveraged ETF will display a 3x change in daily price over the regular fund. If volatility is measured as standard deviation (which I believe is the case), then multiplying a random variable's values by 3 will produce a random variable with 3 times the standard deviation.

History and backtests could always show something different. But to first order, this is approximately what should happen no?
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 8:57 pm

You're missing the purpose of TMF in the strategy as a whole. The volatility is a positive: combined with the negative correlation, the volatility is meant to serve as insurance for an equity crash. TMF delivers slightly more return than EDV during normal times, and a significantly higher rebalancing bonus when equities crash.
Ok so it relies on that negative correlation right? If that's not the case, then you get the result (roughly) I calculated above yes?

I guess that's the part I don't like.
Equity premium? I'd bet money on it.
Term premium? I'd bet money on it.
That a specific correlation stays sufficiently negative between two assets in the future? I would most definitely not bet money on that one.
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 8:57 pm
EDIT: as far as returns go, if you look at annual returns TMF does return about 2x EDV:

https://www.portfoliovisualizer.com/fun ... hlyReturns
Before I reference anything from that graph: the above shows a period of falling interest rates. Since we don't know if they will go up or down, I think the fair comparison is to assume they stay constant. The ER seems soo high, while the term premium so low, that it doesn't seem to me like a good idea.

This is what gw and I keep trying to tell you. We keep saying "but we're not sure this would do well at all if rates stayed constant or went up. It only looks good because rates have been going down".

To which you respond with back test examples of periods with rates going down? Come on man throw us a bone here.

Like gw asked you:
gw wrote:
Tue Jun 11, 2019 5:05 pm
But what are your expected returns going forward, if say, rates stagnate as you suggested? What if they just return to the levels of ten years ago? Is that so implausible? What is your best-case scenario for further long term rate declines, and how does it compare to the alternatives? It just feels like you're not seriously considering these issues.
Is this something that has been tackled on this thread? It's a long way so just point me to the page/poster if so (if not too much trouble!).
1. TMF is not supposed to be 3x EDV, it's supposed to be 3x VGLT. EDV is longer duration than VGLT; that's why TMF is only 2x EDV.
2. Why would falling rates affect the fact that TMF is 2x EDV? It's 2x EDV on the upside and the downside. So if rates rose TMF would be 2x EDV on the downside. The point of my graph that you refused to look at is that TMF is actually tracking the annual performance of an unleveraged bond fund relatively as expected.
3. The only negative correlation scenario that I'm asking you to trust is the flight-to-safety effect: when equities crash, money flows into long Treasuries. I'd say you are safe betting money on that specific scenario.
4. The expense ratio simply doesn't matter over a sufficiently long timeframe. The average CAGR for the strategy over 10 year rolling periods is something like 20%. Take 1% out of that number, would you even notice?
5. Find me some time periods when long rates stayed flat, and I'm happy to run the charts for you.

User avatar
privatefarmer
Posts: 504
Joined: Mon Sep 08, 2014 2:45 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by privatefarmer » Tue Jun 11, 2019 11:13 pm

This thread has so many comments and has been so informative now for the last several months that I’m having trouble keeping up with all the data. Question more for siamond/hedgefundie: at some point you had to rely on monthly data to create your simulated 3x returns as opposed to daily data, correct? Because the daily data wasn’t available? At what point in the backrest was this?

I am wondering because I have found several periods since 1983 where long term yields have either gone up or stayed flat over several years (for example June 2003-sept 2018) where this strategy performed very well. So I am wondering if part of the reason why this strategy did so poorly from 1966-1982 is that the data was not as available to simulate the backtest with? Or is there another reason? Could it be that inflation was out of control many of those years and thus borrowing costs were just too much of a drag? I don’t think it can simply be pegged to rising rates as there definitely have been periods of rising rates since 1983 where this strategy has done well.

Anyhow would love any insight thanks.

MotoTrojan
Posts: 4528
Joined: Wed Feb 01, 2017 8:39 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by MotoTrojan » Tue Jun 11, 2019 11:19 pm

privatefarmer wrote:
Tue Jun 11, 2019 11:13 pm
This thread has so many comments and has been so informative now for the last several months that I’m having trouble keeping up with all the data. Question more for siamond/hedgefundie: at some point you had to rely on monthly data to create your simulated 3x returns as opposed to daily data, correct? Because the daily data wasn’t available? At what point in the backrest was this?

I am wondering because I have found several periods since 1983 where long term yields have either gone up or stayed flat over several years (for example June 2003-sept 2018) where this strategy performed very well. So I am wondering if part of the reason why this strategy did so poorly from 1966-1982 is that the data was not as available to simulate the backtest with? Or is there another reason? Could it be that inflation was out of control many of those years and thus borrowing costs were just too much of a drag? I don’t think it can simply be pegged to rising rates as there definitely have been periods of rising rates since 1983 where this strategy has done well.

Anyhow would love any insight thanks.
2003-2018 looks like a smooth drop in yields to me.

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Tue Jun 11, 2019 11:21 pm

gw wrote:
Tue Jun 11, 2019 9:34 pm
In your longest available backtest, it has twice the std dev of 100% equities, a 1.5x worse max drawdown (75%), and essentially the same returns (10% vs. 11%), so a pretty big fail.

But worse than that, the strategy essentially *only* made money during the post-1981 period of steadily falling rates - it steadily *lost* money throughout the previous period, when rates were rising.

https://i.imgur.com/9D8QjKf.png

Your response is to simply discard the (long!) period where your strategy underperformed, while keeping all the returns from the period where it outperformed, under an argument that rising rates are a thing of the past.
I am the one who pulled out those two time periods and built the graphs: viewtopic.php?f=10&t=272007&start=1050#p4426381 I'm not discarding anything.

Let me quote myself from 1000 posts ago:
So with this new data, I think the bottom line for the strategy is this:

If you believe we will ever return to the Federal Reserve interest rate regime of pre-1982, then you should probably not pursue this.

If, however, you believe that the Fed post-Volcker is a fundamentally different place, reliant on market forces to influence interest rates, then this strategy is a great way to get rich over the long term.
And let me just add, if we do end up forgetting all the lessons of Volcker and lose control of inflation, much of our established understanding of the market and daily life would be upended. This strategy "failing" would probably be the least of your concerns.

User avatar
305pelusa
Posts: 227
Joined: Fri Nov 16, 2018 10:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by 305pelusa » Tue Jun 11, 2019 11:53 pm

HEDGEFUNDIE wrote:
Tue Jun 11, 2019 11:05 pm
305pelusa wrote:
Tue Jun 11, 2019 9:29 pm
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 8:57 pm

It's actually only doubling the volatility of EDV, but anyway...
Could you explain how that's the case? A 3x leveraged ETF will display a 3x change in daily price over the regular fund. If volatility is measured as standard deviation (which I believe is the case), then multiplying a random variable's values by 3 will produce a random variable with 3 times the standard deviation.

History and backtests could always show something different. But to first order, this is approximately what should happen no?
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 8:57 pm

You're missing the purpose of TMF in the strategy as a whole. The volatility is a positive: combined with the negative correlation, the volatility is meant to serve as insurance for an equity crash. TMF delivers slightly more return than EDV during normal times, and a significantly higher rebalancing bonus when equities crash.
Ok so it relies on that negative correlation right? If that's not the case, then you get the result (roughly) I calculated above yes?

I guess that's the part I don't like.
Equity premium? I'd bet money on it.
Term premium? I'd bet money on it.
That a specific correlation stays sufficiently negative between two assets in the future? I would most definitely not bet money on that one.
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 8:57 pm
EDIT: as far as returns go, if you look at annual returns TMF does return about 2x EDV:

https://www.portfoliovisualizer.com/fun ... hlyReturns
Before I reference anything from that graph: the above shows a period of falling interest rates. Since we don't know if they will go up or down, I think the fair comparison is to assume they stay constant. The ER seems soo high, while the term premium so low, that it doesn't seem to me like a good idea.

This is what gw and I keep trying to tell you. We keep saying "but we're not sure this would do well at all if rates stayed constant or went up. It only looks good because rates have been going down".

To which you respond with back test examples of periods with rates going down? Come on man throw us a bone here.

Like gw asked you:
gw wrote:
Tue Jun 11, 2019 5:05 pm
But what are your expected returns going forward, if say, rates stagnate as you suggested? What if they just return to the levels of ten years ago? Is that so implausible? What is your best-case scenario for further long term rate declines, and how does it compare to the alternatives? It just feels like you're not seriously considering these issues.
Is this something that has been tackled on this thread? It's a long way so just point me to the page/poster if so (if not too much trouble!).
1. TMF is not supposed to be 3x EDV, it's supposed to be 3x VGLT. EDV is longer duration than VGLT; that's why TMF is only 2x EDV.
2. Why would falling rates affect the fact that TMF is 2x EDV? It's 2x EDV on the upside and the downside. So if rates rose TMF would be 2x EDV on the downside. The point of my graph that you refused to look at is that TMF is actually tracking the annual performance of an unleveraged bond fund relatively as expected.
3. The only negative correlation scenario that I'm asking you to trust is the flight-to-safety effect: when equities crash, money flows into long Treasuries. I'd say you are safe betting money on that specific scenario.
4. The expense ratio simply doesn't matter over a sufficiently long timeframe. The average CAGR for the strategy over 10 year rolling periods is something like 20%. Take 1% out of that number, would you even notice?
5. Find me some time periods when long rates stayed flat, and I'm happy to run the charts for you.
1. Whoops my bad. Had my durations mixed. Just assumed I meant VGLT every time I said EVD. So TMF would give 3x the VGLT volatility with only 35% additional gain. Is that fair?

2. If rates fall, then the VGLT (and correspondingly TMF) gains are dominated by capital gains, which overshadow the high expense ratio. So it's not surprising that when you look at periods of falling rates, you see TMF tracking 3x its benchmark (or x2 the EDV benchmark). I'm thinking like (very simplified):

If VGLT returns 6%, then TMF returns 17% (15%-1%). That's 14/5 = 2.83. Pretty close to three times the returns.
If VGLT returns 1%, then TMF returns 2% (3%-1%). That's twice the returns.
If VGLT returns 0.5%, then TMF also returns that. Still three times the volatility though!


I haven't thought about the mathematical ramifications but I don't think it's correct to analyze the performance of TMF if rates stay the same by looking at periods where they fall. On the downside, the ER is felt much more.

I didn't refuse to look at the graph. I just don't think it offers insight into the comparison I'm trying to make.

3. Yes, that's a scenario I am not willing to bet on personally. I'll bet on premiums because they expose me to risks (which should be compensated). The flight-to-safety effect is not a premium in that sense.

4. Read point 2. We are in agreement that historically, rates have been constantly falling, giving the bonds amazing capital gain returns that make the ERs basically irrelevant. If rates don't drop, then that's a different story.

5. I was just interested if you had analyzed these scenarios (flat or rising rates). Mathematically, or with Monte Carlos, or whatever else.I think that was gw's point too. Based on the fact that you replied with a backtest of falling rates and are offering to run data I provide, would it be fair to say that these have not been analyzed much?

I don't have data for you, sorry.

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Wed Jun 12, 2019 12:11 am

305pelusa wrote:
Tue Jun 11, 2019 11:53 pm

5. I was just interested if you had analyzed these scenarios (flat or rising rates). Mathematically, or with Monte Carlos, or whatever else.I think that was gw's point too. Based on the fact that you replied with a backtest of falling rates and are offering to run data I provide, would it be fair to say that these have not been analyzed much?
And you would be wrong. The original post includes two Monte Carlo analyses, one including the 1968-1981 rising rate period. If you search for privatefarmer there are numerous posts by him/her backtesting periods of rising/flat rates. Others as well.

User avatar
coingaroo
Posts: 14
Joined: Fri Apr 26, 2019 11:31 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by coingaroo » Wed Jun 12, 2019 4:51 am

HEDGEFUNDIE wrote:
Tue Jun 11, 2019 9:38 pm
coingaroo wrote:
Tue Jun 11, 2019 9:23 pm

Your strategy is analogous to saying a 40/60 stock and long term treasury allocation have fared well in the past with low drawdowns, meaning you can 3x it, giving you roughly 2x the returns with 3x the risk and drawdowns.

That’s not a good trade, and 60% long term treasuries will certainly do poorly in some regimes.
Take a look at the original backtest, over 30 years it delivered 7x the returns, 2x the risk, and 1x the max drawdown. Find me a better trade than that.
If you truly want to make a leveraged,regime independent portfolio, you should either look into adding trend following futures to the mix.

Managed futures (trend following) exhibit a ‘smile’ and generate positive returns during stock market crashes, and positive returns over the long run, independent of interest rates movements.

A 3x leveraged, 35% stock, 35% long term treasuries, 20% managed futures allocation would fare substantially better historically — and expected to win in a rising interest rate environment too. You can access this as a retail investor via IB portfolio margin.
Interesting idea, feel free to start your own thread. My suspicion, however, is that you’ll run into the same buzzsaw of criticism that I have over the course of the last 2000 posts. “Managed futures are opaque derivatives that deserve no place in a BH portfolio” so on and so forth...
You can’t eat backtest returns, just like how you can’t select mutual funds based on past performance.

Don’t get me wrong, I believe in your overall premise of taking a low volatility portfolio, and using leverage to reintroduce volatility and end up with amplified returns and a higher sharpe.

It is an entirely valid concept, and it works due to portable beta, traditional aversion to leverage, and quite simply more risk taking. It’s been buzzing quite a bit around the tactical investing circles: https://econompicdata.blogspot.com/2019 ... y.html?m=1

I’m just not convinced of the specific method of how you have created this low volatility core (ie 40/60), and I think you’re using a number of concepts like risk premia and risk parity incorrectly.

Your strategy mathematically cannot derive a _premia_ from term because your ETF costs outweigh it; your strategy returns come from a relatively expensive bet on treasury yields.

Of course, I understand you’re a passionate fan of this strategy, and that’s cool. Investing is a world with many different opinions and beliefs, and time will tell.

gw
Posts: 756
Joined: Thu May 28, 2009 10:12 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by gw » Wed Jun 12, 2019 7:19 am

HEDGEFUNDIE wrote:
Wed Jun 12, 2019 12:11 am
305pelusa wrote:
Tue Jun 11, 2019 11:53 pm

5. I was just interested if you had analyzed these scenarios (flat or rising rates). Mathematically, or with Monte Carlos, or whatever else.I think that was gw's point too. Based on the fact that you replied with a backtest of falling rates and are offering to run data I provide, would it be fair to say that these have not been analyzed much?
And you would be wrong. The original post includes two Monte Carlo analyses, one including the 1968-1981 rising rate period. If you search for privatefarmer there are numerous posts by him/her backtesting periods of rising/flat rates. Others as well.
I notice your strategy looks pretty bad in the 1968-2018 Monte Carlo you cite, with worse median 20-year performance than the sp500 and something like 40% of 20-year returns below zero (vs. ~7% of them for sp500), despite sampling only one out of five decades with rising rates.

User avatar
305pelusa
Posts: 227
Joined: Fri Nov 16, 2018 10:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by 305pelusa » Wed Jun 12, 2019 7:25 am

HEDGEFUNDIE wrote:
Wed Jun 12, 2019 12:11 am
305pelusa wrote:
Tue Jun 11, 2019 11:53 pm

5. I was just interested if you had analyzed these scenarios (flat or rising rates). Mathematically, or with Monte Carlos, or whatever else.I think that was gw's point too. Based on the fact that you replied with a backtest of falling rates and are offering to run data I provide, would it be fair to say that these have not been analyzed much?
And you would be wrong. The original post includes two Monte Carlo analyses, one including the 1968-1981 rising rate period. If you search for privatefarmer there are numerous posts by him/her backtesting periods of rising/flat rates. Others as well.
No comments to my other points?

Also bear with me (this is what I gather). The 1968-1981 period was brutal to the strategy, so much so that you recommend not using the portfolio if you expect it to repeat itself.

So the math seems to say flat or rising rates would be bad. The backtest with rising rates certainly shows it was bad. So gw and I keep saying the amazing returns you expect of 20+% are dominated by the constantly decreasing rates, not as much by the underlying premia you mentioned. Hence, if rates don't keep going down, the strategy will underperform its benchmark.
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 5:17 pm
Rising interest rates are good for the bond portion of the strategy in the long run and bad for the bond portion of the strategy in the short run. How it affects equities is anyone’s guess. With stagnant interest rates the bonds will do just fine, with the yield to maturity staying constant and levered up.

But as you say, predicting rates is an uninteresting exercise. I’m much more interested in the underlying risk premia that make the strategy tick.
Yet it seems to me like the strategy relies on rates constantly going down from here on out. Also, equities better go up with rising rates. That's fundamental to the strategy as far as I understand.

Any ways, I'm certainly not the first claiming that the results are distorted by the bond bull market in this thread. I guess my point boils down to:
This strategy is relying extensively on rates going down from here on out and/or the negative correlation between the two funds. Neither one of things I would bet money on. Risk premia? Yes. Negative correlations or rates doing X? I would not.

That's all. Excited to see how it keeps going though :)

User avatar
market timer
Posts: 6121
Joined: Tue Aug 21, 2007 1:42 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by market timer » Wed Jun 12, 2019 7:38 am

Speaking of term premia, the Kim-Wright model of the term premium is near an all-time low:

Image

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Wed Jun 12, 2019 8:16 am

gw wrote:
Wed Jun 12, 2019 7:19 am
HEDGEFUNDIE wrote:
Wed Jun 12, 2019 12:11 am
305pelusa wrote:
Tue Jun 11, 2019 11:53 pm

5. I was just interested if you had analyzed these scenarios (flat or rising rates). Mathematically, or with Monte Carlos, or whatever else.I think that was gw's point too. Based on the fact that you replied with a backtest of falling rates and are offering to run data I provide, would it be fair to say that these have not been analyzed much?
And you would be wrong. The original post includes two Monte Carlo analyses, one including the 1968-1981 rising rate period. If you search for privatefarmer there are numerous posts by him/her backtesting periods of rising/flat rates. Others as well.
I notice your strategy looks pretty bad in the 1968-2018 Monte Carlo you cite, with worse median 20-year performance than the sp500 and something like 40% of 20-year returns below zero (vs. ~7% of them for sp500), despite sampling only one out of five decades with rising rates.
If by “pretty bad” you mean a positively skewed distribution with the negative outcomes limited to -20% CAGR and positive outcomes reaching up to +40% CAGR, then yes, just terrible.

User avatar
Kevin M
Posts: 10805
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by Kevin M » Wed Jun 12, 2019 9:08 am

305pelusa wrote:
Tue Jun 11, 2019 7:36 pm
I'm with gw on this one. LIBOR rate is at 2.41%. Current long term treasury yields are at 2.875%. That's a term premium of 0.465%.
First, where are you getting 2.875% from? Yield on 20-year Treasury is 2.42%, and 30-year is 2.62% (as of 6/11), and those are up 8-9 basis points from the beginning of the month.

Next, the term premium is not the difference between a long rate and short rate, but the difference between a long rate and the expected average of a short rate over the long time period. In other words, the difference between say holding a 10-year Treasury to maturity and rolling 1-month Treasuries for 10 years.

So to determine a term premium you need a way to estimate the future course of the short-term rate. Of course no one is good at predicting interest rates, so there's a lot of uncertainty in term premium estimates.

As I shared in an earlier post, the term premium of the 10-year Treasury is about -0.9% according to one respected model (the model does not go beyond 10-year maturity). Someone else just shared results of another model that also puts the 10-year Treasury in negative territory.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
305pelusa
Posts: 227
Joined: Fri Nov 16, 2018 10:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by 305pelusa » Wed Jun 12, 2019 9:37 am

Kevin M wrote:
Wed Jun 12, 2019 9:08 am
305pelusa wrote:
Tue Jun 11, 2019 7:36 pm
I'm with gw on this one. LIBOR rate is at 2.41%. Current long term treasury yields are at 2.875%. That's a term premium of 0.465%.
First, where are you getting 2.875% from? Yield on 20-year Treasury is 2.42%, and 30-year is 2.62% (as of 6/11), and those are up 8-9 basis points from the beginning of the month.

Next, the term premium is not the difference between a long rate and short rate, but the difference between a long rate and the expected average of a short rate over the long time period. In other words, the difference between say holding a 10-year Treasury to maturity and rolling 1-month Treasuries for 10 years.

So to determine a term premium you need a way to estimate the future course of the short-term rate. Of course no one is good at predicting interest rates, so there's a lot of uncertainty in term premium estimates.

As I shared in an earlier post, the term premium of the 10-year Treasury is about -0.9% according to one respected model (the model does not go beyond 10-year maturity). Someone else just shared results of another model that also puts the 10-year Treasury in negative territory.

Kevin
Huh you're right. The treasury yield is much lower than I realized. I must've been looking at the wrong stats in TreasuryDirect. This makes the strategy even worse as of now if rates stay the same.

The term premium TMF captures is the difference between what it pays to borrow (below LIBOR) and what it invests in (2.42 as you mentioned). I'm assuming the LIBOR rate stays constant in my calculation.

So for 100 dollars invested, you get:
100*0.0242+ 200*(0.0242-0.0183) -100 = 2.6

So a 2.6% return if rates (long and LIBOR) stay stagnant, like Hedgefundie mentioned vs just 2.42% return by buying the treasury itself. So you get an increase of 7% return for 3 times the volatility.

It sounds like a mistake I'm making is assuming the LIBOR rate stays constant. If it drops, it does better. If it increases it does worse. I don't really know what will occur so I just take it as is now but if you're expecting the term premium to be 0.9, then clearly you expect the LIBOR rate to drop. Would that be a fair statement?

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Wed Jun 12, 2019 9:52 am

coingaroo wrote:
Wed Jun 12, 2019 4:51 am
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 9:38 pm
coingaroo wrote:
Tue Jun 11, 2019 9:23 pm

Your strategy is analogous to saying a 40/60 stock and long term treasury allocation have fared well in the past with low drawdowns, meaning you can 3x it, giving you roughly 2x the returns with 3x the risk and drawdowns.

That’s not a good trade, and 60% long term treasuries will certainly do poorly in some regimes.
Take a look at the original backtest, over 30 years it delivered 7x the returns, 2x the risk, and 1x the max drawdown. Find me a better trade than that.
If you truly want to make a leveraged,regime independent portfolio, you should either look into adding trend following futures to the mix.

Managed futures (trend following) exhibit a ‘smile’ and generate positive returns during stock market crashes, and positive returns over the long run, independent of interest rates movements.

A 3x leveraged, 35% stock, 35% long term treasuries, 20% managed futures allocation would fare substantially better historically — and expected to win in a rising interest rate environment too. You can access this as a retail investor via IB portfolio margin.
Interesting idea, feel free to start your own thread. My suspicion, however, is that you’ll run into the same buzzsaw of criticism that I have over the course of the last 2000 posts. “Managed futures are opaque derivatives that deserve no place in a BH portfolio” so on and so forth...
You can’t eat backtest returns, just like how you can’t select mutual funds based on past performance.

Don’t get me wrong, I believe in your overall premise of taking a low volatility portfolio, and using leverage to reintroduce volatility and end up with amplified returns and a higher sharpe.

It is an entirely valid concept, and it works due to portable beta, traditional aversion to leverage, and quite simply more risk taking. It’s been buzzing quite a bit around the tactical investing circles: https://econompicdata.blogspot.com/2019 ... y.html?m=1

I’m just not convinced of the specific method of how you have created this low volatility core (ie 40/60), and I think you’re using a number of concepts like risk premia and risk parity incorrectly.
Whether I am using the terms incorrectly or not is uninteresting to me, what I care about is whether it makes money for me.

It's interesting that so many people have suggested improvements to my strategy to protect against rising rates, but no one has actually started their own thread and subjected themselves to criticism over 2000 posts like I have.

I suspect that is because if you were to actually construct such an alternative strategy (with managed futures, gold, or what have you), your backtest would show great performance in the 60s and 70s, and then significant underperformance from the 80s onward.

In short, you don't get inflation protection for free - the tradeoff you make during low inflation times is "wasted" portfolio space that could be riding the stock / bond bull. This is the same reason why the Permanent Portfolio attracts a ton of interest but relatively little money.

So between these two options: my portfolio which has done great over the past 30 years, or this alternative portfolio which would have been a laggard for the past 30 years, which one would you honestly put your money on?

MotoTrojan
Posts: 4528
Joined: Wed Feb 01, 2017 8:39 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by MotoTrojan » Wed Jun 12, 2019 9:56 am

305pelusa wrote:
Wed Jun 12, 2019 9:37 am

Huh you're right. The treasury yield is much lower than I realized. I must've been looking at the wrong stats in TreasuryDirect. This makes the strategy even worse as of now if rates stay the same.

The term premium TMF captures is the difference between what it pays to borrow (below LIBOR) and what it invests in (2.42 as you mentioned). I'm assuming the LIBOR rate stays constant in my calculation.

So for 100 dollars invested, you get:
100*0.0242+ 200*(0.0242-0.0183) -100 = 2.6

So a 2.6% return if rates (long and LIBOR) stay stagnant, like Hedgefundie mentioned vs just 2.42% return by buying the treasury itself. So you get an increase of 7% return for 3 times the volatility.

It sounds like a mistake I'm making is assuming the LIBOR rate stays constant. If it drops, it does better. If it increases it does worse. I don't really know what will occur so I just take it as is now but if you're expecting the term premium to be 0.9, then clearly you expect the LIBOR rate to drop. Would that be a fair statement?
As a skeptical participant (but with a healthy dose for good measure) I appreciate this analysis, but I think you are focusing too much on the absolute return generated by TMF and not looking at the full portfolio.

To make this easier to share I am using actual funds so the timeframe is short, but it still proves my point. In this bull market UPRO has performed well on it's own, and achieves a CAGR of almost 29% but with almost 39% volatility. The portfolio "only" achieves 27.5% CAGR, but knocks volatility down to just 21%. TMF alone though only achieved 14% CAGR with almost 39% volatility. So even though TMF achieved less than half the CAGR of UPRO, it barely reduced overall performance even as the majority holding at 60% allocation, and nearly halved the volatility.

https://www.portfoliovisualizer.com/bac ... total3=100

When extended back to 1982 you get the following:
UPRO 13% CAGR 45% STD
TMF 14% CAGR 35% STD
Portfolio 20% CAGR 28% STD

1955-now:
UPRO 10.5% CAGR 44% STD
TMF 2.3% CAGR 31% STD
Portfolio 10.4% CAGR 26% STD

1955-1982:
UPRO 7.6% CAGR 43% STD
TMF -8.9% CAGR 25% STD
Portfolio 0.4% CAGR 24.5% STD

In all cases it performed significantly better than the weighted average of return/volatility.

MotoTrojan
Posts: 4528
Joined: Wed Feb 01, 2017 8:39 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by MotoTrojan » Wed Jun 12, 2019 10:11 am

HEDGEFUNDIE wrote:
Tue Jun 11, 2019 7:40 pm


Here is something you are missing - the financing spread on TMF is actually negative, i.e. the financing rate is below LIBOR. You can confirm this via the latest annual reports.

When interest rates are expected to rise, it makes sense that TMF can get paid to take on that risk via swaps.
HEDGEFUNDIE, could you remind me the assumption made in the simulated data on financing spread? I believe it was left out entirely (making the results conservative if we assume negative spreads are common)? Can't recall if a further understanding of historical spreads or reasonable assumptions was ever made beyond just picking a number.

User avatar
305pelusa
Posts: 227
Joined: Fri Nov 16, 2018 10:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by 305pelusa » Wed Jun 12, 2019 10:19 am

MotoTrojan wrote:
Wed Jun 12, 2019 9:56 am
305pelusa wrote:
Wed Jun 12, 2019 9:37 am

Huh you're right. The treasury yield is much lower than I realized. I must've been looking at the wrong stats in TreasuryDirect. This makes the strategy even worse as of now if rates stay the same.

The term premium TMF captures is the difference between what it pays to borrow (below LIBOR) and what it invests in (2.42 as you mentioned). I'm assuming the LIBOR rate stays constant in my calculation.

So for 100 dollars invested, you get:
100*0.0242+ 200*(0.0242-0.0183) -100 = 2.6

So a 2.6% return if rates (long and LIBOR) stay stagnant, like Hedgefundie mentioned vs just 2.42% return by buying the treasury itself. So you get an increase of 7% return for 3 times the volatility.

It sounds like a mistake I'm making is assuming the LIBOR rate stays constant. If it drops, it does better. If it increases it does worse. I don't really know what will occur so I just take it as is now but if you're expecting the term premium to be 0.9, then clearly you expect the LIBOR rate to drop. Would that be a fair statement?
As a skeptical participant (but with a healthy dose for good measure) I appreciate this analysis, but I think you are focusing too much on the absolute return generated by TMF and not looking at the full portfolio.

To make this easier to share I am using actual funds so the timeframe is short, but it still proves my point. In this bull market UPRO has performed well on it's own, and achieves a CAGR of almost 29% but with almost 39% volatility. The portfolio "only" achieves 27.5% CAGR, but knocks volatility down to just 21%. TMF alone though only achieved 14% CAGR with almost 39% volatility. So even though TMF achieved less than half the CAGR of UPRO, it barely reduced overall performance even as the majority holding at 60% allocation, and nearly halved the volatility.

https://www.portfoliovisualizer.com/bac ... total3=100

When extended back to 1982 you get the following:
UPRO 13% CAGR 45% STD
TMF 14% CAGR 35% STD
Portfolio 20% CAGR 28% STD

1955-now:
UPRO 10.5% CAGR 44% STD
TMF 2.3% CAGR 31% STD
Portfolio 10.4% CAGR 26% STD

1955-1982:
UPRO 7.6% CAGR 43% STD
TMF -8.9% CAGR 25% STD
Portfolio 0.4% CAGR 24.5% STD

In all cases it performed significantly better than the weighted average of return/volatility.
Haha yes you're very right. I am only focusing on TMF returns.

I rationalize that because I don't really feel strongly about that flight-to-safety effect and the negative correlation. Yes historically it's been negative. And yes I sort of buy the argument that it should be negative. But not enough to put money that relies on it.

The potential to get that correlation wrong due to survivorship bias and data mining feels really high to me.

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Wed Jun 12, 2019 10:24 am

MotoTrojan wrote:
Wed Jun 12, 2019 10:11 am
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 7:40 pm


Here is something you are missing - the financing spread on TMF is actually negative, i.e. the financing rate is below LIBOR. You can confirm this via the latest annual reports.

When interest rates are expected to rise, it makes sense that TMF can get paid to take on that risk via swaps.
HEDGEFUNDIE, could you remind me the assumption made in the simulated data on financing spread? I believe it was left out entirely (making the results conservative if we assume negative spreads are common)? Can't recall if a further understanding of historical spreads or reasonable assumptions was ever made beyond just picking a number.
Yes it was left out entirely.

MotoTrojan
Posts: 4528
Joined: Wed Feb 01, 2017 8:39 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by MotoTrojan » Wed Jun 12, 2019 10:28 am

305pelusa wrote:
Wed Jun 12, 2019 10:19 am
MotoTrojan wrote:
Wed Jun 12, 2019 9:56 am
305pelusa wrote:
Wed Jun 12, 2019 9:37 am

Huh you're right. The treasury yield is much lower than I realized. I must've been looking at the wrong stats in TreasuryDirect. This makes the strategy even worse as of now if rates stay the same.

The term premium TMF captures is the difference between what it pays to borrow (below LIBOR) and what it invests in (2.42 as you mentioned). I'm assuming the LIBOR rate stays constant in my calculation.

So for 100 dollars invested, you get:
100*0.0242+ 200*(0.0242-0.0183) -100 = 2.6

So a 2.6% return if rates (long and LIBOR) stay stagnant, like Hedgefundie mentioned vs just 2.42% return by buying the treasury itself. So you get an increase of 7% return for 3 times the volatility.

It sounds like a mistake I'm making is assuming the LIBOR rate stays constant. If it drops, it does better. If it increases it does worse. I don't really know what will occur so I just take it as is now but if you're expecting the term premium to be 0.9, then clearly you expect the LIBOR rate to drop. Would that be a fair statement?
As a skeptical participant (but with a healthy dose for good measure) I appreciate this analysis, but I think you are focusing too much on the absolute return generated by TMF and not looking at the full portfolio.

To make this easier to share I am using actual funds so the timeframe is short, but it still proves my point. In this bull market UPRO has performed well on it's own, and achieves a CAGR of almost 29% but with almost 39% volatility. The portfolio "only" achieves 27.5% CAGR, but knocks volatility down to just 21%. TMF alone though only achieved 14% CAGR with almost 39% volatility. So even though TMF achieved less than half the CAGR of UPRO, it barely reduced overall performance even as the majority holding at 60% allocation, and nearly halved the volatility.

https://www.portfoliovisualizer.com/bac ... total3=100

When extended back to 1982 you get the following:
UPRO 13% CAGR 45% STD
TMF 14% CAGR 35% STD
Portfolio 20% CAGR 28% STD

1955-now:
UPRO 10.5% CAGR 44% STD
TMF 2.3% CAGR 31% STD
Portfolio 10.4% CAGR 26% STD

1955-1982:
UPRO 7.6% CAGR 43% STD
TMF -8.9% CAGR 25% STD
Portfolio 0.4% CAGR 24.5% STD

In all cases it performed significantly better than the weighted average of return/volatility.
Haha yes you're very right. I am only focusing on TMF returns.

I rationalize that because I don't really feel strongly about that flight-to-safety effect and the negative correlation. Yes historically it's been negative. And yes I sort of buy the argument that it should be negative. But not enough to put money that relies on it.

The potential to get that correlation wrong due to survivorship bias and data mining feels really high to me.
Fair enough. I see it as a core component of this portfolio’s design so your feelings make sense if you don’t put faith in it, but wanted to point out that it has historically worked without simply riding the wave of TMF alone.

User avatar
Kevin M
Posts: 10805
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by Kevin M » Wed Jun 12, 2019 12:52 pm

305pelusa wrote:
Wed Jun 12, 2019 9:37 am
It sounds like a mistake I'm making is assuming the LIBOR rate stays constant. If it drops, it does better. If it increases it does worse. I don't really know what will occur so I just take it as is now but if you're expecting the term premium to be 0.9, then clearly you expect the LIBOR rate to drop. Would that be a fair statement?
First, the 10-year Treasury term premium I quoted was -0.9% (that's a negative number). That was the month-end number for May. The June 10 number is -0.81%, so it has increased a bit in June.

This is based on the results of a term premium model; the results are published by the Federal Reserve Bank of New York here: https://www.newyorkfed.org/research/dat ... remia.html. The model is referred to as the ACM model, based on the last names of the authors of the paper that develops the model.

The math of the model is way over my head, but a simple explanation of the Treasury term premium is that it's the expected excess annualized return from holding an N-year Treasury compared to rolling a 1-month Treasury monthly for for N years. So the model currently estimates that you'll earn about 0.8% more per year by rolling 1-month Treasuries for 10 years than by holding a 10-year Treasury to maturity.

The model doesn't cover maturities greater than ten years, so it doesn't give us an estimate for the 20-year term premium, which probably would be more relevant to this strategy.

Again, lots of uncertainty in estimates of short-term yields out to 10 years, but at least it's based on economic science rather than just people's opinions.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
305pelusa
Posts: 227
Joined: Fri Nov 16, 2018 10:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by 305pelusa » Wed Jun 12, 2019 1:30 pm

Kevin M wrote:
Wed Jun 12, 2019 12:52 pm
305pelusa wrote:
Wed Jun 12, 2019 9:37 am
It sounds like a mistake I'm making is assuming the LIBOR rate stays constant. If it drops, it does better. If it increases it does worse. I don't really know what will occur so I just take it as is now but if you're expecting the term premium to be 0.9, then clearly you expect the LIBOR rate to drop. Would that be a fair statement?
First, the 10-year Treasury term premium I quoted was -0.9% (that's a negative number). That was the month-end number for May. The June 10 number is -0.81%, so it has increased a bit in June.

This is based on the results of a term premium model; the results are published by the Federal Reserve Bank of New York here: https://www.newyorkfed.org/research/dat ... remia.html. The model is referred to as the ACM model, based on the last names of the authors of the paper that develops the model.

The math of the model is way over my head, but a simple explanation of the Treasury term premium is that it's the expected excess annualized return from holding an N-year Treasury compared to rolling a 1-month Treasury monthly for for N years. So the model currently estimates that you'll earn about 0.8% more per year by rolling 1-month Treasuries for 10 years than by holding a 10-year Treasury to maturity.

The model doesn't cover maturities greater than ten years, so it doesn't give us an estimate for the 20-year term premium, which probably would be more relevant to this strategy.

Again, lots of uncertainty in estimates of short-term yields out to 10 years, but at least it's based on economic science rather than just people's opinions.

Kevin
I literally have no idea what the ST treasury rate has to do with any of this.

User avatar
Kevin M
Posts: 10805
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by Kevin M » Wed Jun 12, 2019 3:19 pm

305pelusa wrote:
Wed Jun 12, 2019 1:30 pm
I literally have no idea what the ST treasury rate has to do with any of this.
You used the term "term premium", but you used it incorrectly. I'm simply clarifying what "term premium" means in the fixed-income world.

Google "term premium".

First link I get: https://www.bloomberg.com/news/articles ... cktake-q-a. Just from the little bit I can read without a subscription:
1. What exactly is a term premium, anyway?

It’s the difference between what you get for locking up your money for an extended period and what you would get if you simply kept rolling over short-term instruments for the same amount of time.
(underline mine)

Which is basically what I explained. When discussing the Treasury term premium, the "short-term instrument" is a short-term Treasury, typically the shortest term for which rates are published, which is one month.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
305pelusa
Posts: 227
Joined: Fri Nov 16, 2018 10:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by 305pelusa » Wed Jun 12, 2019 4:34 pm

Kevin M wrote:
Wed Jun 12, 2019 3:19 pm
305pelusa wrote:
Wed Jun 12, 2019 1:30 pm
I literally have no idea what the ST treasury rate has to do with any of this.
You used the term "term premium", but you used it incorrectly. I'm simply clarifying what "term premium" means in the fixed-income world.

Google "term premium".

First link I get: https://www.bloomberg.com/news/articles ... cktake-q-a. Just from the little bit I can read without a subscription:
1. What exactly is a term premium, anyway?

It’s the difference between what you get for locking up your money for an extended period and what you would get if you simply kept rolling over short-term instruments for the same amount of time.
(underline mine)

Which is basically what I explained. When discussing the Treasury term premium, the "short-term instrument" is a short-term Treasury, typically the shortest term for which rates are published, which is one month.

Kevin
I know that colloquially it refers to the premium of ST vs LT treasuries. That's not what the OP meant. He means the term premium in the context of TMF (borrowing at the short term LIBOR rate or below and investing in LT treasuries). I understood the context of his post and answered in those terms.

But I get how it could be confusing to you to use ambiguous terms so I can refer to it as "TMF-specific term premium" or something like that from now on :)

klaus14
Posts: 168
Joined: Sun Nov 25, 2018 7:43 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by klaus14 » Wed Jun 12, 2019 5:18 pm

klaus14 wrote:
Fri Feb 15, 2019 10:55 pm
NTSX is a cheaper way to achieve some risk parity.
It invests 90% in SP500 and 10% is the collateral for treasury contracts to achieve 60% exposure.

If you do 50% NTSX and 50% EDV, It translates to 45% SP500 and 80% treasuries.

It is less leveraged but cost is considerably small. NTSX expense ratio is 0.20% and EDV is 0.07%.
FYI this cheap/mild version of this strategy also performed well as expected. It has better sharpe/sortino and less drawdowns. If you look at monthly returns, directions are same but magnitude is scaled.

NTSX is new, but here is a comparison with "simulated" NTSX (VFITX+SPY-CASHX). Again better sharpe/sortino compared to balanced fund and 3x

User avatar
Kevin M
Posts: 10805
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by Kevin M » Wed Jun 12, 2019 6:45 pm

305pelusa wrote:
Wed Jun 12, 2019 4:34 pm
Kevin M wrote:
Wed Jun 12, 2019 3:19 pm
305pelusa wrote:
Wed Jun 12, 2019 1:30 pm
I literally have no idea what the ST treasury rate has to do with any of this.
You used the term "term premium", but you used it incorrectly. I'm simply clarifying what "term premium" means in the fixed-income world.
<snip>
I know that colloquially it refers to the premium of ST vs LT treasuries.
Not sure "colloquial" is the right term:
col·lo·qui·al
/kəˈlōkwēəl/
adjective
(of language) used in ordinary or familiar conversation; not formal or literary.
(underline mine)

"Term premium" is used by academics in papers they write, so most definitely a formal term.
305pelusa wrote:
Wed Jun 12, 2019 4:34 pm
That's not what the OP meant. He means the term premium in the context of TMF (borrowing at the short term LIBOR rate or below and investing in LT treasuries). I understood the context of his post and answered in those terms.
You mean this post?
HEDGEFUNDIE wrote:
Tue Jun 11, 2019 10:32 am
GW’s objection raises the question: what is it reasonable to expect ex-ante? I submit the following:
<snip>
2. That the term premium will persist.
<snip>
These are risk-based premia that comprise the fundamental drivers of the financial markets.
<snip>
I don't read your interpretation into the actual words that were written. I don't know what OP meant, but referring to "risk-based premia that comprise the fundamental drivers of the financial markets" makes it sound to me like we should be interpreting term premium as its standard meaning.

I actually responded directly to that post, pointing out that the term premium on the 10-year Treasury is negative--somewhat as a counterpoint to "the term premium will persist". I don't recall OP responding directly to that point, which would help clarify the intended meaning. I suggest we stick to the standard meaning of "term premium" and other commonly used investing/financial terms in our investing/financial discussions.
305pelusa wrote:
Wed Jun 12, 2019 4:34 pm
But I get how it could be confusing to you to use ambiguous terms so I can refer to it as "TMF-specific term premium" or something like that from now on :)
It's not confusing, just incorrect. "Term premium" has a widely understood meaning in finance--at least that's my impression from seeing it used many times. I don't recall ever seeing it used as the difference between a yield and a borrowing rate to achieve that yield.

If OP did indeed mean a risk premium for borrowing short and lending long, then it would be good to clarify that. That's certainly not what I interpret as a "term premium", although it is a type of risk premium.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
305pelusa
Posts: 227
Joined: Fri Nov 16, 2018 10:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by 305pelusa » Wed Jun 12, 2019 7:01 pm

Kevin M wrote:
Wed Jun 12, 2019 6:45 pm

Not sure "colloquial" is the right term:
Probably not.
Kevin M wrote:
Wed Jun 12, 2019 6:45 pm
You mean this post?
Yep.
Kevin M wrote:
Wed Jun 12, 2019 6:45 pm

I don't read your interpretation into the actual words that were written. I don't know what OP meant, but referring to "risk-based premia that comprise the fundamental drivers of the financial markets" makes it sound to me like we should be interpreting term premium as its standard meaning.
That's ok. I'm using some context clues there which might or might not be write (see below).
Kevin M wrote:
Wed Jun 12, 2019 6:45 pm

I actually responded directly to that post, pointing out that the term premium on the 10-year Treasury is negative--somewhat as a counterpoint to "the term premium will persist". I don't recall OP responding directly to that point, which would help clarify the intended meaning.
He probably didn't answer because you were talking about something different than he was.
Kevin M wrote:
Wed Jun 12, 2019 6:45 pm
I suggest we stick to the standard meaning of "term premium" and other commonly used investing/financial terms in our investing/financial discussions.
I concur.
Kevin M wrote:
Wed Jun 12, 2019 6:45 pm
It's not confusing, just incorrect.
Well I think it did lead to confusion. We can agree to disagree.
Kevin M wrote:
Wed Jun 12, 2019 6:45 pm
If OP did indeed mean a risk premium for borrowing short and lending long, then it would be good to clarify that. That's certainly not what I interpret as a "term premium", although it is a type of risk premium.

Kevin
I know it's not what you interpreted but I can almost bet anything I got his meaning right. Mostly from that post where I figured out that the term premium in the context of TMF would be the LT rate minus LIBOR and he corrected that TMF actually borrows below LIBOR. And that hence, the premium would be better. So I think he's referring to a type of term premium (specific to TMF) but not necessarily THE term premium.

But like I said before, I agree with you man. I'll be more specific with the wording. Perhaps OP will chime in what what he meant. Let's just leave this behind all right?

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Wed Jun 12, 2019 7:47 pm

Term premium: borrowing short to buy long, or in other words the delta between the short rate and the long rate. The other details don’t really matter (LIBOR is going away soon, presumably to be replaced by a short Treasury rate, and the financing spread could be positive or negative, so I’m assuming the long run spread is zero).

I didn’t respond to Kevin M because the idea that the currently inverted yield curve could become a long run condition is so preposterous as to merit no comment.

To take it seriously is tantamount to overturning the laws of supply & demand, attributing negative value to liquidity, and other crazy notions.

petulant
Posts: 492
Joined: Thu Sep 22, 2016 1:09 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by petulant » Wed Jun 12, 2019 7:57 pm

HEDGEFUNDIE,

My apologies for not wading through 40 pages, but could you direct me to any part of the discussion that discusses pros/cons to attempting to construct this strategy from derivatives directly rather than leveraged ETFs?

Thanks,

petulant

Topic Author
HEDGEFUNDIE
Posts: 2581
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Wed Jun 12, 2019 7:58 pm

petulant wrote:
Wed Jun 12, 2019 7:57 pm
HEDGEFUNDIE,

My apologies for not wading through 40 pages, but could you direct me to any part of the discussion that discusses pros/cons to attempting to construct this strategy from derivatives directly rather than leveraged ETFs?

Thanks,

petulant
Either search for “futures” or PM hdas.

petulant
Posts: 492
Joined: Thu Sep 22, 2016 1:09 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by petulant » Thu Jun 13, 2019 6:23 am

HEDGEFUNDIE wrote:
Wed Jun 12, 2019 7:58 pm
petulant wrote:
Wed Jun 12, 2019 7:57 pm
HEDGEFUNDIE,

My apologies for not wading through 40 pages, but could you direct me to any part of the discussion that discusses pros/cons to attempting to construct this strategy from derivatives directly rather than leveraged ETFs?

Thanks,

petulant
Either search for “futures” or PM hdas.
Has anybody discussed using options?

Staying the Course
Posts: 12
Joined: Wed Dec 26, 2018 12:31 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by Staying the Course » Thu Jun 13, 2019 1:16 pm

How much does this strategy suffer in a taxable account? Would the cumulative taxes/fees mitigate any advantage over, say 100% VTSAX?

(I prefer the simplicity of keeping my core retirement portfolio intact and with Vanguard.)

sarabayo
Posts: 265
Joined: Fri Jun 29, 2018 6:59 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by sarabayo » Thu Jun 13, 2019 1:42 pm

Staying the Course wrote:
Thu Jun 13, 2019 1:16 pm
How much does this strategy suffer in a taxable account? Would the cumulative taxes/fees mitigate any advantage over, say 100% VTSAX?

(I prefer the simplicity of keeping my core retirement portfolio intact and with Vanguard.)
The trailing 12 month distribution yield of UPRO is 0.63%, and TMF's is 1.33%. That means that HEDGEFUNDIE's strategy has a total distribution yield of 1.05%. In comparison, VTI has a distribution yield of 2.00% (and presumably so does VTSAX, thanks to Vanguard's ETF share class patent). So it would seem that the strategy is actually more tax-efficient than 100% VTSAX (?).

See also the discussion on page 29 of this thread: viewtopic.php?f=10&t=272007&p=4457252#p4457252

Of course, the above doesn't account for capital gains incurred when rebalancing; a 100% VTSAX doesn't need to rebalance at all, so 100% VTSAX definitely holds the advantage there.

MotoTrojan
Posts: 4528
Joined: Wed Feb 01, 2017 8:39 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by MotoTrojan » Thu Jun 13, 2019 1:49 pm

sarabayo wrote:
Thu Jun 13, 2019 1:42 pm
Staying the Course wrote:
Thu Jun 13, 2019 1:16 pm
How much does this strategy suffer in a taxable account? Would the cumulative taxes/fees mitigate any advantage over, say 100% VTSAX?

(I prefer the simplicity of keeping my core retirement portfolio intact and with Vanguard.)
The trailing 12 month distribution yield of UPRO is 0.63%, and TMF's is 1.33%. That means that HEDGEFUNDIE's strategy has a total distribution yield of 1.05%. In comparison, VTI has a distribution yield of 2.00% (and presumably so does VTSAX, thanks to Vanguard's ETF share class patent). So it would seem that the strategy is actually more tax-efficient than 100% VTSAX (?).

See also the discussion on page 29 of this thread: viewtopic.php?f=10&t=272007&p=4457252#p4457252

Of course, the above doesn't account for capital gains incurred when rebalancing; a 100% VTSAX doesn't need to rebalance at all, so 100% VTSAX definitely holds the advantage there.
The allocation is more tax-efficient than 100% VTSAX but the strategy relies on frequent rebalancing, so I would argue the strategy itself is not more tax-efficient. Also TMF is not going to be a qualified dividend so directly comparing 1:1 with the distribution yield of VTI is flawed.

Post Reply