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

[MoneyMarathon]

It's pegged to +/- 2 years from Bloomberg Barclays Long-Term Government/Credit Index. So it's roughly 12.5 to 16.5.

Thank you. Skimmed some of the docs but couldn't find the target exposure. 100% equity and XX% bond?

I think they allow themselves to exceed 100% bond by a little (currently it looks like 115% or so), which helps them get close to their secondary index, the Bloomberg Barclays Long-Term Government/Credit Index, after expenses (but before taxes, of course):

https://www.bloomberg.com/quote/BFALTRUU:IND

The level to which they end up tracking the index, net of expenses, has been pretty good.



Their target is to meet or beat the secondary index (before expenses).

We all know that's super hard, but they also passively track the S&P 500 and limit their strategies to a duration target, so the net result is consistently pretty great until stocks take a dive (or the return of stagflation).

This secondary index seems to be the same index tracked by Vanguard's BLV, making this fund basically a leveraged VOO/BLV, with an active aspect to the bond side:

https://investor.vanguard.com/etf/profile/BLV

[MotoTrojan]

It's pegged to +/- 2 years from Bloomberg Barclays Long-Term Government/Credit Index. So it's roughly 12.5 to 16.5.

Thank you. Skimmed some of the docs but couldn't find the target exposure. 100% equity and XX% bond?

I think they allow themselves to exceed 100% bond by a little (currently it looks like 115% or so), which helps them get close to their secondary index after expenses (but before taxes, of course):

https://www.bloomberg.com/quote/BFALTRUU:IND

The level to which they end up tracking the index, net of expenses, has been pretty good.

And 100% equity? So this would be equivalent to 100/115 or so? Not far from risk-parity it seems..

[MoneyMarathon]

And 100% equity? So this would be equivalent to 100/115 or so? Not far from risk-parity it seems..

Given the level of duration on the bond side (around 14 instead of around 24 like EDV or around 55 like TMF), which reduces the volatility of their bonds, it's got a lot less volatility from the bond side.

I believe risk parity is about 40% stock and 60% bonds here, so I guess you're right, it's not that far after all. Mostly less leveraged, so less volatile overall, less duration risk, and a greater allocation to credit risk factors (instead of just duration on the bond side).

[MotoTrojan]

And 100% equity? So this would be equivalent to 100/115 or so? Not far from risk-parity it seems..

Given the level of duration on the bond side (around 14 instead of around 24 like EDV or around 55 like TMF), which reduces the volatility of their bonds, it's got a lot less volatility from the bond side.

I believe risk parity is about 40% stock and 60% bonds here, so I guess you're right, it's not that far after all. Mostly less leveraged, so less volatile overall, less duration risk, and a greater allocation to credit risk factors (instead of just duration on the bond side).

Makes sense. Has done well overall but seems the reduced bond exposure and higher credit risk prevented much improvement in drawdown. Still an interesting fund, good luck!

[MoneyMarathon]

the reduced bond exposure and higher credit risk prevented much improvement in drawdown

Yes, but I'm pretty happy with the reduced duration exposure because I don't know what the next bear market looks like.

[butricksaid]

1. There is no intellectual justification for the quarterly rebalance vs other time periods, so there is even less justification for choosing calendar quarterly vs “whenever you start quarterly”

2. Don’t market time. Just don’t.

So does this mean I need to wait for my quarterly schedule to make additional, on-going contributions?

For those of us who opt for adding more funds, say with our bi-weekly or monthly paychecks (in addition to the lump-sum initial contribution), would we need to wait until a quarterly rebalancing event? Does the rebalance strategy change to accommodate the extra funds or do the contributions workaround the rebalancing?

If you contribution and rebalance at the same time, presumably the order would be to make your contribution first then rebalance to minimize capital gains.

[vel]

If we enumerate the set of all possible undesirable final outcomes regarding the TMF/UPRO strategy, putting aside assessments of likelihood, perhaps we have:
1) They both get wiped out.
2) One of the two gets wiped out.
3) They both lose value in such a way and to such an extent that recovery is extremely unlikely in the operable future.
4) One of the two falls into (3).

OP's response "the strategy would be over" to the scenario of correlation between stocks and long treasuries increasing (esp. during downturns) raises the question - is the intention to ride it out until it terminates in one of cases {1,2,3,4}, or to identify that it has transitioned into a scenario that should trigger an exit from the strategy?

If there are cases where the detection of a particular scenario should trigger an exit, and would provide enough time to do so, I think it would be helpful to try to lay out and formalize these contingencies.

While we have some information regarding possible risks and some comments on their likelihood, I think we may benefit from teasing apart the scenarios and trying to understand which are actionable.

Also, perhaps some quantification for example, regarding a threshold signaling the correlation between stocks and long treasuries having increased to such an extent as to cause the strategy to be rendered unworkable.

[HEDGEFUNDIE]

If we enumerate the set of all possible undesirable final outcomes regarding the TMF/UPRO strategy, putting aside assessments of likelihood, perhaps we have:
1) They both get wiped out.
2) One of the two gets wiped out.
3) They both lose value in such a way and to such an extent that recovery is extremely unlikely in the operable future.
4) One of the two falls into (3).

OP's response "the strategy would be over" to the scenario of correlation between stocks and long treasuries increasing (esp. during downturns) raises the question - is the intention to ride it out until it terminates in one of cases {1,2,3,4}, or to identify that it has transitioned into a scenario that should trigger an exit from the strategy?

If there are cases where the detection of a particular scenario should trigger an exit, and would provide enough time to do so, I think it would be helpful to try to lay out and formalize these contingencies.

While we have some information regarding possible risks and some comments on their likelihood, I think we may benefit from teasing apart the scenarios and trying to understand which are actionable.

Also, perhaps some quantification for example, regarding a threshold signaling the correlation between stocks and long treasuries having increased to such an extent as to cause the strategy to be rendered unworkable.

Interesting question. The last time the correlation between the S&P and long Treasuries was generally positive was in 1986-1997. The correlation was 0.30 during this period, compared to -0.33 afterward.

During 1986-1997, the strategy returned 25% CAGR.

[vel]

I am interpreting this as the basis for a point that is unclear.

[schismal]

FWIW I settled on the UPRO 1-month look-back volatility balance against a 50/50 TMF/EDV bond-side. Will be a wild ride and I look forward to it!

I just plugged in the numbers for July so far, and the UPRO/TMF ratio came out to 60/40. Curious if you plan to cap the allotment of either of those assets, or if you just plan to rebalance monthly however the lookback dictates.

Edit: UPRO and 50/50 TMF/EDV comes out to ~54/(23/23), but same question.

[MotoTrojan]

1. There is no intellectual justification for the quarterly rebalance vs other time periods, so there is even less justification for choosing calendar quarterly vs “whenever you start quarterly”

2. Don’t market time. Just don’t.

So does this mean I need to wait for my quarterly schedule to make additional, on-going contributions?

For those of us who opt for adding more funds, say with our bi-weekly or monthly paychecks (in addition to the lump-sum initial contribution), would we need to wait until a quarterly rebalancing event? Does the rebalance strategy change to accommodate the extra funds or do the contributions workaround the rebalancing?

If you contribution and rebalance at the same time, presumably the order would be to make your contribution first then rebalance to minimize capital gains.

Personally I’d never touch this strategy in a taxable account.

I am rebalancing monthly now but when I intended to use quarterly I was making regular contributions using M1 and would just let it buy the underweight asset. I see no reason to wait until the quarter.

[MotoTrojan]

FWIW I settled on the UPRO 1-month look-back volatility balance against a 50/50 TMF/EDV bond-side. Will be a wild ride and I look forward to it!

I just plugged in the numbers for July so far, and the UPRO/TMF ratio came out to 60/40. Curious if you plan to cap the allotment of either of those assets, or if you just plan to rebalance monthly however the lookback dictates.

Edit: UPRO and 50/50 TMF/EDV comes out to ~54/(23/23), but same question.

The 50/50 bond split reduces the equity extremes as you see here but my plan was to follow it strictly.

[caklim00]

FWIW I settled on the UPRO 1-month look-back volatility balance against a 50/50 TMF/EDV bond-side. Will be a wild ride and I look forward to it!

I just plugged in the numbers for July so far, and the UPRO/TMF ratio came out to 60/40. Curious if you plan to cap the allotment of either of those assets, or if you just plan to rebalance monthly however the lookback dictates.

Edit: UPRO and 50/50 TMF/EDV comes out to ~54/(23/23), but same question.

The 50/50 bond split reduces the equity extremes as you see here but my plan was to follow it strictly.

I'm still waiting on M1 to approve my account (they sure take awhile since my credit bureaus were frozen and they didn't auto approve). What exactly is the proposal for someone using 50/50 TMF/EDV on the bond side? Is it to analyze what the split should be between UPRO and Bonds on a monthly basis? Am I reading that correctly that for right now that would equate to 46% UPRO, 23% TMF, 23% EDV?

[MotoTrojan]

FWIW I settled on the UPRO 1-month look-back volatility balance against a 50/50 TMF/EDV bond-side. Will be a wild ride and I look forward to it!

I just plugged in the numbers for July so far, and the UPRO/TMF ratio came out to 60/40. Curious if you plan to cap the allotment of either of those assets, or if you just plan to rebalance monthly however the lookback dictates.

Edit: UPRO and 50/50 TMF/EDV comes out to ~54/(23/23), but same question.

The 50/50 bond split reduces the equity extremes as you see here but my plan was to follow it strictly.

I'm still waiting on M1 to approve my account (they sure take awhile since my credit bureaus were frozen and they didn't auto approve). What exactly is the proposal for someone using 50/50 TMF/EDV on the bond side? Is it to analyze what the split should be between UPRO and Bonds on a monthly basis? Am I reading that correctly that for right now that would equate to 46% UPRO, 23% TMF, 23% EDV?

54 on UPRO. The strategy is simply to reset your allocation based on the past months volatility. So when UPRO has been calm like now you’ll hold more. If volatility spikes you’ll decrease. The 50-50 bond side is a separate impact which I’m including to reduce bond volatility some, in effect reducing leverage modestly and also reducing the extent of the UPRO swings. Historically it would still have some months of 70%+ UPRO every now and again.

You could do this with just UPRO and EDV for example but then your longterm average UPRO exposure would be closer to 30%.

[caklim00]

FWIW I settled on the UPRO 1-month look-back volatility balance against a 50/50 TMF/EDV bond-side. Will be a wild ride and I look forward to it!

I just plugged in the numbers for July so far, and the UPRO/TMF ratio came out to 60/40. Curious if you plan to cap the allotment of either of those assets, or if you just plan to rebalance monthly however the lookback dictates.

Edit: UPRO and 50/50 TMF/EDV comes out to ~54/(23/23), but same question.

The 50/50 bond split reduces the equity extremes as you see here but my plan was to follow it strictly.

I'm still waiting on M1 to approve my account (they sure take awhile since my credit bureaus were frozen and they didn't auto approve). What exactly is the proposal for someone using 50/50 TMF/EDV on the bond side? Is it to analyze what the split should be between UPRO and Bonds on a monthly basis? Am I reading that correctly that for right now that would equate to 46% UPRO, 23% TMF, 23% EDV?

54 on UPRO. The strategy is simply to reset your allocation based on the past months volatility. So when UPRO has been calm like now you’ll hold more. If volatility spikes you’ll decrease. The 50-50 bond side is a separate impact which I’m including to reduce bond volatility some, in effect reducing leverage modestly and also reducing the extent of the UPRO swings. Historically it would still have some months of 70%+ UPRO every now and again.

You could do this with just UPRO and EDV for example but then your longterm average UPRO exposure would be closer to 30%.

Interesting idea. Do you have a formula you plug this in to spit out the UPRO percentage? You just doing this at the beginning of the month looking back at the previous month?

EDIT: I'm trying to play around with portfolio visulizer to figure out how you came up with 56/23/23 right now and am struggling

[MotoTrojan]

I just plugged in the numbers for July so far, and the UPRO/TMF ratio came out to 60/40. Curious if you plan to cap the allotment of either of those assets, or if you just plan to rebalance monthly however the lookback dictates.

Edit: UPRO and 50/50 TMF/EDV comes out to ~54/(23/23), but same question.

The 50/50 bond split reduces the equity extremes as you see here but my plan was to follow it strictly.

I'm still waiting on M1 to approve my account (they sure take awhile since my credit bureaus were frozen and they didn't auto approve). What exactly is the proposal for someone using 50/50 TMF/EDV on the bond side? Is it to analyze what the split should be between UPRO and Bonds on a monthly basis? Am I reading that correctly that for right now that would equate to 46% UPRO, 23% TMF, 23% EDV?

54 on UPRO. The strategy is simply to reset your allocation based on the past months volatility. So when UPRO has been calm like now you’ll hold more. If volatility spikes you’ll decrease. The 50-50 bond side is a separate impact which I’m including to reduce bond volatility some, in effect reducing leverage modestly and also reducing the extent of the UPRO swings. Historically it would still have some months of 70%+ UPRO every now and again.

You could do this with just UPRO and EDV for example but then your longterm average UPRO exposure would be closer to 30%.

Interesting idea. Do you have a formula you plug this in to spit out the UPRO percentage? You just doing this at the beginning of the month looking back at the previous month?

EDIT: I'm trying to play around with portfolio visulizer to figure out how you came up with 56/23/23 right now and am struggling

PV will do this for you with one bond fund but it’s simple math. Allocation of UPRO = (1/upro_vol)/((1/upro_vol)+(1/bond_vol)). For me the bond vol is simply the average of both funds since their correlation is essentially 1.

I trade last day of month but that’s arbitrary. Vol comes from past 20 trading days.

[caklim00]

PV will do this for you with one bond fund but it’s simple math. Allocation of UPRO = (1/upro_vol)/((1/upro_vol)+(1/bond_vol)). For me the bond vol is simply the average of both funds since their correlation is essentially 1.

I trade last day of month but that’s arbitrary. Vol comes from past 20 trading days.

Sorry for being such a pain. Where do you find the volume in PV? Or ar you just clicking into one of the sections? I was trying to use the Portfolio Optimization section using Risk Parity as Optimization Goal, but realize I am way off.

I'm trying to setup an easy way that I can calculate every month else I'm better off trying a static allocation such as 50/50 and then just doing quarterly rebalancing.

[MoneyMarathon]

I'm better off trying a static allocation such as 50/50 and then just doing quarterly rebalancing.

Good idea. Calculation leads to thinking, thinking leads to doubt, doubt leads to fear, fear leads to the dark side.

[MotoTrojan]

I'm better off trying a static allocation such as 50/50 and then just doing quarterly rebalancing.

Good idea. Calculation leads to thinking, thinking leads to doubt, doubt leads to fear, fear leads to the dark side.

Agreed, you'll need some strong will to stick with calculation. Having a 75% UPRO allocation when the market talk is all about 50% declines will take some will.

I would do a lot more research on volatility managed portfolios before blindly following my reckless plan .

[MotoTrojan]

PV will do this for you with one bond fund but it’s simple math. Allocation of UPRO = (1/upro_vol)/((1/upro_vol)+(1/bond_vol)). For me the bond vol is simply the average of both funds since their correlation is essentially 1.

I trade last day of month but that’s arbitrary. Vol comes from past 20 trading days.

Sorry for being such a pain. Where do you find the volume in PV? Or ar you just clicking into one of the sections? I was trying to use the Portfolio Optimization section using Risk Parity as Optimization Goal, but realize I am way off.

I'm trying to setup an easy way that I can calculate every month else I'm better off trying a static allocation such as 50/50 and then just doing quarterly rebalancing.

Step by step:
1: PV website, click on Adaptive Allocation under Timing Models (bottom right)
2: Add your tickers (only use 2 for now)
3: Set Volatility Period to Specify and then manually enter 20 days
4: Monthly rebalance
5: Run the test
6: Under results, click Timing Periods tab
7: Scroll to the very bottom and it'll show you the allocation based on the current signal date, as well as the raw volatility of each fund

If you wanted to adapt my 50/50 bond strategy or anything similar then you could add all three assets, turn the number of assets to 3, and then use the some process but you would ignore the allocation recommendation and just use PV as a tool to give you the raw volatility of each holding, then do the math yourself.

Equity markets are getting awfully calm right now. Will be an interesting rebalance on Wednesday!

[rascott]

PV will do this for you with one bond fund but it’s simple math. Allocation of UPRO = (1/upro_vol)/((1/upro_vol)+(1/bond_vol)). For me the bond vol is simply the average of both funds since their correlation is essentially 1.

I trade last day of month but that’s arbitrary. Vol comes from past 20 trading days.

Sorry for being such a pain. Where do you find the volume in PV? Or ar you just clicking into one of the sections? I was trying to use the Portfolio Optimization section using Risk Parity as Optimization Goal, but realize I am way off.

I'm trying to setup an easy way that I can calculate every month else I'm better off trying a static allocation such as 50/50 and then just doing quarterly rebalancing.

Step by step:
1: PV website, click on Adaptive Allocation under Timing Models (bottom right)
2: Add your tickers (only use 2 for now)
3: Set Volatility Period to Specify and then manually enter 20 days
4: Monthly rebalance
5: Run the test
6: Under results, click Timing Periods tab
7: Scroll to the very bottom and it'll show you the allocation based on the current signal date, as well as the raw volatility of each fund

If you wanted to adapt my 50/50 bond strategy or anything similar then you could add all three assets, turn the number of assets to 3, and then use the some process but you would ignore the allocation recommendation and just use PV as a tool to give you the raw volatility of each holding, then do the math yourself.

Equity markets are getting awfully calm right now. Will be an interesting rebalance on Wednesday!

True.... what I'm getting is current signal is roughly 60/40 UPRO/TMF. Is that accurate?

I'm sure it's buried in this thread somewhere....but how did you come up with 20 day look back being the best option?

[caklim00]

PV will do this for you with one bond fund but it’s simple math. Allocation of UPRO = (1/upro_vol)/((1/upro_vol)+(1/bond_vol)). For me the bond vol is simply the average of both funds since their correlation is essentially 1.

I trade last day of month but that’s arbitrary. Vol comes from past 20 trading days.

Sorry for being such a pain. Where do you find the volume in PV? Or ar you just clicking into one of the sections? I was trying to use the Portfolio Optimization section using Risk Parity as Optimization Goal, but realize I am way off.

I'm trying to setup an easy way that I can calculate every month else I'm better off trying a static allocation such as 50/50 and then just doing quarterly rebalancing.

Step by step:
1: PV website, click on Adaptive Allocation under Timing Models (bottom right)
2: Add your tickers (only use 2 for now)
3: Set Volatility Period to Specify and then manually enter 20 days
4: Monthly rebalance
5: Run the test
6: Under results, click Timing Periods tab
7: Scroll to the very bottom and it'll show you the allocation based on the current signal date, as well as the raw volatility of each fund

If you wanted to adapt my 50/50 bond strategy or anything similar then you could add all three assets, turn the number of assets to 3, and then use the some process but you would ignore the allocation recommendation and just use PV as a tool to give you the raw volatility of each holding, then do the math yourself.

Equity markets are getting awfully calm right now. Will be an interesting rebalance on Wednesday!

Here's what I get when I run that today:
Signal Date Assets Details
07/24/2019 40.27% Direxion Daily 20+ Yr Trsy Bull 3X ETF (TMF)
59.73% ProShares UltraPro S&P500 (UPRO) Next period predicted allocation based on market data as of 07/24/2019.
UPRO performance over 6-month window is 46.55%.
UPRO volatility over 20-day volatility window is 1.32%.
TMF performance over 6-month window is 30.80%.
TMF volatility over 20-day volatility window is 1.96%.

[MotoTrojan]

PV will do this for you with one bond fund but it’s simple math. Allocation of UPRO = (1/upro_vol)/((1/upro_vol)+(1/bond_vol)). For me the bond vol is simply the average of both funds since their correlation is essentially 1.

I trade last day of month but that’s arbitrary. Vol comes from past 20 trading days.

Sorry for being such a pain. Where do you find the volume in PV? Or ar you just clicking into one of the sections? I was trying to use the Portfolio Optimization section using Risk Parity as Optimization Goal, but realize I am way off.

I'm trying to setup an easy way that I can calculate every month else I'm better off trying a static allocation such as 50/50 and then just doing quarterly rebalancing.

Step by step:
1: PV website, click on Adaptive Allocation under Timing Models (bottom right)
2: Add your tickers (only use 2 for now)
3: Set Volatility Period to Specify and then manually enter 20 days
4: Monthly rebalance
5: Run the test
6: Under results, click Timing Periods tab
7: Scroll to the very bottom and it'll show you the allocation based on the current signal date, as well as the raw volatility of each fund

If you wanted to adapt my 50/50 bond strategy or anything similar then you could add all three assets, turn the number of assets to 3, and then use the some process but you would ignore the allocation recommendation and just use PV as a tool to give you the raw volatility of each holding, then do the math yourself.

Equity markets are getting awfully calm right now. Will be an interesting rebalance on Wednesday!

True.... what I'm getting is current signal is roughly 60/40 UPRO/TMF. Is that accurate?

I'm sure it's buried in this thread somewhere....but how did you come up with 20 day look back being the best option?

Backtesting. Also hydromod found that going even shorter only improved things but you’d start bumping up against trading cost issues.

It’s a trade off either way. Longer smooths things out for less extremes but doesn’t react as fast to a market change.

20 days is ~1 month of trading days.

[noraz123]

Step by step:
1: PV website, click on Adaptive Allocation under Timing Models (bottom right)
2: Add your tickers (only use 2 for now)
3: Set Volatility Period to Specify and then manually enter 20 days
4: Monthly rebalance
5: Run the test
6: Under results, click Timing Periods tab
7: Scroll to the very bottom and it'll show you the allocation based on the current signal date, as well as the raw volatility of each fund

Thank you so much for this! I was trying to figure this out myself. While I feel quite comfortable with all the math, I did not want to manually calculate all of this, and PV website is a little intimidating for the uninitiated.

[noraz123]

Equity markets are getting awfully calm right now. Will be an interesting rebalance on Wednesday!

To confirm - volatility of TMF is currently higher than UPRO. Using numbers from today (UPRO Vol = 1.32%, TMF Vol = 1.96%), the weightings are basically reversed! 59.8% UPRO and 40.2% TMF.

Is this correct?

And thanks again for all the help explaining. Very much appreciated.

[MotoTrojan]

Equity markets are getting awfully calm right now. Will be an interesting rebalance on Wednesday!

To confirm - volatility of TMF is currently higher than UPRO. Using numbers from today (UPRO Vol = 1.32%, TMF Vol = 1.96%), the weightings are basically reversed! 59.8% UPRO and 40.2% TMF.

Is this correct?

And thanks again for all the help explaining. Very much appreciated.

Looks right to me. If you run a backtest you'll find times in the past where it was >80% UPRO.

[mickens16]

Equity markets are getting awfully calm right now. Will be an interesting rebalance on Wednesday!

To confirm - volatility of TMF is currently higher than UPRO. Using numbers from today (UPRO Vol = 1.32%, TMF Vol = 1.96%), the weightings are basically reversed! 59.8% UPRO and 40.2% TMF.

Is this correct?

And thanks again for all the help explaining. Very much appreciated.

Looks right to me. If you run a backtest you'll find times in the past where it was >80% UPRO.

You're now splitting TMF/EDV (50/50), would your allocation be 60% UPRO, 20% TMF, 20% EDV? Or are you running the program with 3 etf's to get the lookback?

[MotoTrojan]

Equity markets are getting awfully calm right now. Will be an interesting rebalance on Wednesday!

To confirm - volatility of TMF is currently higher than UPRO. Using numbers from today (UPRO Vol = 1.32%, TMF Vol = 1.96%), the weightings are basically reversed! 59.8% UPRO and 40.2% TMF.

Is this correct?

And thanks again for all the help explaining. Very much appreciated.

Looks right to me. If you run a backtest you'll find times in the past where it was >80% UPRO.

You're now splitting TMF/EDV (50/50), would your allocation be 60% UPRO, 20% TMF, 20% EDV? Or are you running the program with 3 etf's to get the lookback?

You can run the program with 3 ETFs to get the volatilities, but you'll have to run the eqn yourself to get the allocation. If you run 3 assets the program will split up risk between those 3 so you'd end up with mostly EDV, then some UPRO, then even less TMF.

Because TMF & EDV are nearly perfectly correlated I am taking their volatilities and simply averaging them, then balancing UPRO against that. I haven't run the latest numbers but it would bring the allocation down from 60% UPRO to closer to 52%, and then split the remaining 48% between TMF & EDV.

60/20/20 UPRO/TMF/EDV would not be a risk-parity balance and would be skewing risk towards equities.

[schismal]

Sorry for being such a pain. Where do you find the volume in PV? Or ar you just clicking into one of the sections? I was trying to use the Portfolio Optimization section using Risk Parity as Optimization Goal, but realize I am way off.

I'm trying to setup an easy way that I can calculate every month else I'm better off trying a static allocation such as 50/50 and then just doing quarterly rebalancing.

If you want, just use this spreadsheet: LINK

Go to Yahoo, look up the closing prices of each fund over the last month, and paste into columns B, D, and F. It should spit out the ratio at the bottom.

[MotoTrojan]

Sorry for being such a pain. Where do you find the volume in PV? Or ar you just clicking into one of the sections? I was trying to use the Portfolio Optimization section using Risk Parity as Optimization Goal, but realize I am way off.

I'm trying to setup an easy way that I can calculate every month else I'm better off trying a static allocation such as 50/50 and then just doing quarterly rebalancing.

If you want, just use this spreadsheet: LINK

Go to Yahoo, look up the closing prices of each fund over the last month, and paste into columns B, D, and F. It should spit out the ratio at the bottom.

Some of the papers suggested using an exponential weighting factor to the look-back period (weights more recent days higher); makes sense intuitively and would be quite easy in excel . Looking up the prices for even the last 20 days of three assets each month sounds like a pain though...

[Steve Reading]

To the people using recent-volatility (like previous 20 days) to dynamically shift your allocations, please answer the following:

It is well-documented that market daily returns are not normally distributed. Not even close. The St. Dev. (as a measure of risk) is quite literally a meaningless number for a non-normal distribution.

Over longer horizons (years), the Central Limit Theorem dictates that the return distributions will be approximately normal (even though it is constituted by non-normal daily distributions... that's where the CLT comes in). So over long time horizons, St. Dev does start to make sense. The OP's choice of using long-term volatility is on the right track and theoretically sound IMO.

Knowing this, what makes you think it is prudent in any way to use such a short-term window of standard deviation?

And no "it backtested well!" answers. I'm looking for an actual theoretical or intuitive explanation as to why you think it makes sense.

Thanks.

[rascott]

To the people using recent-volatility (like previous 20 days) to dynamically shift your allocations, please answer the following:

It is well-documented that market daily returns are not normally distributed. Not even close. The St. Dev. (as a measure of risk) is quite literally a meaningless number for a non-normal distribution.

Over longer horizons (years), the Central Limit Theorem dictates that the return distributions will be approximately normal (even though it is constituted by non-normal daily distributions... that's where the CLT comes in). So over long time horizons, St. Dev does start to make sense. The OP's choice of using long-term volatility is on the right track and theoretically sound IMO.

Knowing this, what makes you think it is prudent in any way to use such a short-term window of standard deviation?

And no "it backtested well!" answers. I'm looking for an actual theoretical or intuitive explanation as to why you think it makes sense.

Thanks.

I've had a hard time getting behind that as well, from a theoretical level. I have a finance background educationally, but have never worked in the field. So still learning (re-learning) a lot of the stuff.

I don't put money into something I don't understand.....I understand the OP's model. The short-term volatility adjustment seems to back test well (better)...but I'm not understating why...or if there should be any expectation of it going forward.

[MotoTrojan]

To the people using recent-volatility (like previous 20 days) to dynamically shift your allocations, please answer the following:

It is well-documented that market daily returns are not normally distributed. Not even close. The St. Dev. (as a measure of risk) is quite literally a meaningless number for a non-normal distribution.

Over longer horizons (years), the Central Limit Theorem dictates that the return distributions will be approximately normal (even though it is constituted by non-normal daily distributions... that's where the CLT comes in). So over long time horizons, St. Dev does start to make sense. The OP's choice of using long-term volatility is on the right track and theoretically sound IMO.

Knowing this, what makes you think it is prudent in any way to use such a short-term window of standard deviation?

And no "it backtested well!" answers. I'm looking for an actual theoretical or intuitive explanation as to why you think it makes sense.

Thanks.

https://papers.ssrn.com/sol3/papers.cfm ... id=2659431

There are a few groups researching (and funds implementing) target volatility portfolios. These target a specific volatility on the equity side and then move into other assets (such as long bonds) to meet that target. This is basically what I am doing but instead of having to pick my own target volatility, I am just using risk-parity between both assets.

The studies show that while recent returns do not predict future returns well, recent volatility does. Even if StDev isn't a perfect measure or risk, it is a good measure of volatility, no?

Would love to hear more of your thoughts. I have given myself until the end of the month to truly lockdown my plan. Yes, this is clearly a market-timing strategy just like trend-following is.

[rascott]

Curious as to if anyone has decided to change their asset allocation based on all the possible alternatives MoneyMarathon has provided

I've actually made my changes. Unlike others in this thread, I haven't started a play money or side bet account. I was able to buy PSLDX in the Roth IRAs at Vanguard and in my 401k by using a Schwab PCRA brokerage window. The fees were $20 x 2 at Vanguard and $49.95 x 1 at Schwab.

I decided that I really like PSLDX (S&P 500 and long-term treasury/corporate/other bonds) in tax-advantaged. Part of the reason I like it is behavioral. It generally hugs the S&P 500 index with some of what might be called "alpha" from using bonds, which is ideal for someone who tends to be happy when not underperforming the market. Another behavioral aspect is that there is no button-pushing. This is huge, we're not robots. I didn't even check the financial news at all in 2018 and didn't know there was a stock correction in December until this summer. Ignorance is bliss, I guess. Another reason is that it was easy to explain and get my spouse on board. Having it go through 2008, for real, and be basically fine is another selling point. Being at PIMCO instead of a tiny ETF provider is another selling point. Not having to wonder about optimal rebalancing strategies or forgetting to do it optimally could be considered a plus. Being able to focus more free time and energy on other, income-generating opportunities is a plus. Not having to worry about what would happen if someone inherited the account and left it alone is a plus.

I also decided that I didn't like PISIX, the international fund. They're a bit too active for my tastes on the bond side, whipping around their duration exposure based on manager decisions, I guess because they have no "secondary index" or duration target like PSLDX (that secondary index and duration target forces the PSLDX guys to be sort-of closet indexers at heart, which is good).

The other reason that I don't like PISIX is that PSLDX is too darn good. It takes up valuable tax-advantaged space that is better used by PSLDX, which throws off more income on the bond side. My goal is actually to have 100% of tax advantaged in PSLDX, while keeping international in taxable. I'm not there yet, need to build up my taxable account more. Currently I'm a a little over 80% of tax advantaged in PSLDX, the rest in the institutional version of VXUS (ex-US world equities). Over time, I'd like to shift to using IXUS in taxable (since it seems somewhat more tax-efficient than VXUS) for my international exposure.

Personal Capital says my allocation now looks like this:

-92.8% Cash
13.95% International Bonds
77.7% US Bonds
17.44% International Stock
80.55% US Stock
3.34% Alternatives (Real Estate, etc.)

My paycheck 401k contributions (traditional and after-tax) are going to the international stock fund.

I can't get this fund (PSLDX) out of my head. Have been thinking about it for the last couple of weeks. And trying to figure out why I shouldn't put basically all my LC blend allocation into this fund (or as much as I can get into it).

Under what scenario will this thing greatly lag the SP500? Beyond inflation spikes.... which would be theoretically temporary? And in 'normal' world how could it not over-perform the index?

My only issue right now is that I can't buy it anywhere where I have funds. (TDA will do it for $0 min, but $50 fee, but I have only a tiny account there right now). Other than that... a small Roth at Ally that's my wife's (and has anyone actually confirmed Ally will do it for $10/$100 min?).

Dang, Merrill Edge....come on man. Closed.

I do have brokerage link through Fidelity for both our 401ks.... but think someone mentioned that didn't work either. This is one tough fund to get, lol.


How would one setup a similar position with ETFs? It's obviously not as leveraged as OP (which I'm still doing in a M1 Roth). I still believe futures are the best way to do this type of setup over LETFs...and I'd much rather the pros handle the futures than me.

[firebirdparts]

Okay so I did an experiment today where I bought more PSLDX at Fido, but I already owned some. If I search for it, I can't pull it up. But by starting with what I already own, and clicking the trade button, it allowed me to buy more. FWIW.

[dave_k]

Sorry for being such a pain. Where do you find the volume in PV? Or ar you just clicking into one of the sections? I was trying to use the Portfolio Optimization section using Risk Parity as Optimization Goal, but realize I am way off.

I'm trying to setup an easy way that I can calculate every month else I'm better off trying a static allocation such as 50/50 and then just doing quarterly rebalancing.

If you want, just use this spreadsheet: LINK

Go to Yahoo, look up the closing prices of each fund over the last month, and paste into columns B, D, and F. It should spit out the ratio at the bottom.

Some of the papers suggested using an exponential weighting factor to the look-back period (weights more recent days higher); makes sense intuitively and would be quite easy in excel . Looking up the prices for even the last 20 days of three assets each month sounds like a pain though...

If you're going to do that, would it make sense to update it every day and use bands to rebalance? If the allocation and target drift around in generally the same direction no rebalancing would be needed, but if they diverged quickly then this would respond quickly. I know it would be a pain to check and calculate daily, and the trading costs would vary, but would it be reasonable in theory? Anyone interested in backtesting it? It may be possible to automate grabbing the data daily and giving an alert when a rebalance is called for.

[rascott]

Okay so I did an experiment today where I bought more PSLDX at Fido, but I already owned some. If I search for it, I can't pull it up. But by starting with what I already own, and clicking the trade button, it allowed me to buy more. FWIW.

Guessing after this thread it will get even harder to buy.

[columbia]

Okay so I did an experiment today where I bought more PSLDX at Fido, but I already owned some. If I search for it, I can't pull it up. But by starting with what I already own, and clicking the trade button, it allowed me to buy more. FWIW.

Guessing after this thread it will get even harder to buy.

Dumb move or not: I put my entire VG Roth in it. So have this overall allocation:

40% FSLDX
60% TIAA Traditional + treasuries

[butricksaid]

I want to test this strategy out, I really do, but I can't pull the trigger on it right now. This feels like the wrong time to begin taking this strategy given that we are in the "late stages of a long economic growth cycle".

• Obviously we want to employ this strategy at the bottom of the downturn.
• At best we employ this strategy during the downturn.
• On average we employ this strategy a little bit prior to a downturn.
• At worst we employ this strategy just before the downturn.

[HEDGEFUNDIE]

I want to test this strategy out, I really do, but I can't pull the trigger on it right now. This feels like the wrong time to begin taking this strategy given that we are in the "late stages of a long economic growth cycle".

• Obviously we want to employ this strategy at the bottom of the downturn.
• At best we employ this strategy during the downturn.
• On average we employ this strategy a little bit prior to a downturn.
• At worst we employ this strategy just before the downturn.

None of this market timing matters. If you had started this strategy in Jan 2008, right before the crash, you would be sitting on 18% CAGR gains right now, compared to 8% CAGR for the S&P. And during the crash itself the strategy tracked the S&P pretty closely.

[MotoTrojan]

I want to test this strategy out, I really do, but I can't pull the trigger on it right now. This feels like the wrong time to begin taking this strategy given that we are in the "late stages of a long economic growth cycle".

• Obviously we want to employ this strategy at the bottom of the downturn.
• At best we employ this strategy during the downturn.
• On average we employ this strategy a little bit prior to a downturn.
• At worst we employ this strategy just before the downturn.

What is the alternative investment? If you want to prepare for a potential equity downturn and are already in treasuries or some other low-risk holding state then I can see your point. If you are 100% equity with the funds that would be put into this though, then I don't quite follow. This historically should've outperformed if you had moved 100% S&P500 to this strategy right at the peak of the S&P500.

[tj]

I want to test this strategy out, I really do, but I can't pull the trigger on it right now. This feels like the wrong time to begin taking this strategy given that we are in the "late stages of a long economic growth cycle".

• Obviously we want to employ this strategy at the bottom of the downturn.
• At best we employ this strategy during the downturn.
• On average we employ this strategy a little bit prior to a downturn.
• At worst we employ this strategy just before the downturn.

Not to mention it's gone up 35% this year! Talk about buying high...

[MotoTrojan]

Sorry for being such a pain. Where do you find the volume in PV? Or ar you just clicking into one of the sections? I was trying to use the Portfolio Optimization section using Risk Parity as Optimization Goal, but realize I am way off.

I'm trying to setup an easy way that I can calculate every month else I'm better off trying a static allocation such as 50/50 and then just doing quarterly rebalancing.

If you want, just use this spreadsheet: LINK

Go to Yahoo, look up the closing prices of each fund over the last month, and paste into columns B, D, and F. It should spit out the ratio at the bottom.

Some of the papers suggested using an exponential weighting factor to the look-back period (weights more recent days higher); makes sense intuitively and would be quite easy in excel . Looking up the prices for even the last 20 days of three assets each month sounds like a pain though...

If you're going to do that, would it make sense to update it every day and use bands to rebalance? If the allocation and target drift around in generally the same direction no rebalancing would be needed, but if they diverged quickly then this would respond quickly. I know it would be a pain to check and calculate daily, and the trading costs would vary, but would it be reasonable in theory? Anyone interested in backtesting it? It may be possible to automate grabbing the data daily and giving an alert when a rebalance is called for.

Perhaps, but too much effort for me. The same could apply even if you didn't use an exponential time-weighted volatility.

[MotoTrojan]

I want to test this strategy out, I really do, but I can't pull the trigger on it right now. This feels like the wrong time to begin taking this strategy given that we are in the "late stages of a long economic growth cycle".

• Obviously we want to employ this strategy at the bottom of the downturn.
• At best we employ this strategy during the downturn.
• On average we employ this strategy a little bit prior to a downturn.
• At worst we employ this strategy just before the downturn.

Not to mention it's gone up 35% this year! Talk about buying high...

It has had 100% years, and is now well above those levels. Not sure I follow.

[HEDGEFUNDIE]

I want to test this strategy out, I really do, but I can't pull the trigger on it right now. This feels like the wrong time to begin taking this strategy given that we are in the "late stages of a long economic growth cycle".

• Obviously we want to employ this strategy at the bottom of the downturn.
• At best we employ this strategy during the downturn.
• On average we employ this strategy a little bit prior to a downturn.
• At worst we employ this strategy just before the downturn.

Not to mention it's gone up 35% this year! Talk about buying high...

Over the past 10 years alone, there have been two years when the return has been 70-80% and three other years when the return was 40-50%

[tj]

I want to test this strategy out, I really do, but I can't pull the trigger on it right now. This feels like the wrong time to begin taking this strategy given that we are in the "late stages of a long economic growth cycle".

• Obviously we want to employ this strategy at the bottom of the downturn.
• At best we employ this strategy during the downturn.
• On average we employ this strategy a little bit prior to a downturn.
• At worst we employ this strategy just before the downturn.

Not to mention it's gone up 35% this year! Talk about buying high...

It has had 100% years, and is now well above those levels. Not sure I follow.

The YTD says 35% at https://www.pimco.com/en-us/investments ... -fund/inst

[MotoTrojan]

I want to test this strategy out, I really do, but I can't pull the trigger on it right now. This feels like the wrong time to begin taking this strategy given that we are in the "late stages of a long economic growth cycle".

• Obviously we want to employ this strategy at the bottom of the downturn.
• At best we employ this strategy during the downturn.
• On average we employ this strategy a little bit prior to a downturn.
• At worst we employ this strategy just before the downturn.

Not to mention it's gone up 35% this year! Talk about buying high...

It has had 100% years, and is now well above those levels. Not sure I follow.

The YTD says 35% at https://www.pimco.com/en-us/investments ... -fund/inst

To be clear I was referring to the OP's strategy, not the pimco fund, but both have had great years. The OP's strategy had a best year of 106% in 1995; that was followed up by a totally okay 9% and then two years of ~55% return, each! Since that 106% year it has returned ~18.5% CAGR. Don't market time.

[columbia]

I want to test this strategy out, I really do, but I can't pull the trigger on it right now. This feels like the wrong time to begin taking this strategy given that we are in the "late stages of a long economic growth cycle".

• Obviously we want to employ this strategy at the bottom of the downturn.
• At best we employ this strategy during the downturn.
• On average we employ this strategy a little bit prior to a downturn.
• At worst we employ this strategy just before the downturn.

Not to mention it's gone up 35% this year! Talk about buying high...

It has had 100% years, and is now well above those levels. Not sure I follow.

The YTD says 35% at https://www.pimco.com/en-us/investments ... -fund/inst

To be clear I was referring to the OP's strategy, not the pimco fund, but both have had great years. The OP's strategy had a best year of 106% in 1995; that was followed up by a totally okay 9% and then two years of ~55% return, each! Since that 106% year it has returned ~18.5% CAGR. Don't market time.

The assumption is that those following the strategy of this thread are in it for the long run. You wouldn’t advise a 25 year old to avoid a conventional stock heavy AA.

[tj]

Not to mention it's gone up 35% this year! Talk about buying high...

It has had 100% years, and is now well above those levels. Not sure I follow.

The YTD says 35% at https://www.pimco.com/en-us/investments ... -fund/inst

To be clear I was referring to the OP's strategy, not the pimco fund, but both have had great years. The OP's strategy had a best year of 106% in 1995; that was followed up by a totally okay 9% and then two years of ~55% return, each! Since that 106% year it has returned ~18.5% CAGR. Don't market time.

The assumption is that those following the strategy of this thread are in it for the long run. You wouldn’t advise a 25 year old to avoid a conventional stock heavy AA.

Sure, I just hate the idea of buying high, even if maybe it's irrelevant 30 years from now. Any reaosn I shouldn't sell my Target Date Fund in my Vangard Roth and load up on StocksPlus Long Duration? What's the biggest risk? Clearly it's been a superior performer in the past, but we don't know what will happen in the future.

[dave_k]

Some of the papers suggested using an exponential weighting factor to the look-back period (weights more recent days higher); makes sense intuitively and would be quite easy in excel . Looking up the prices for even the last 20 days of three assets each month sounds like a pain though...

If you're going to do that, would it make sense to update it every day and use bands to rebalance? If the allocation and target drift around in generally the same direction no rebalancing would be needed, but if they diverged quickly then this would respond quickly. I know it would be a pain to check and calculate daily, and the trading costs would vary, but would it be reasonable in theory? Anyone interested in backtesting it? It may be possible to automate grabbing the data daily and giving an alert when a rebalance is called for.

Perhaps, but too much effort for me. The same could apply even if you didn't use an exponential time-weighted volatility.

True, but I figured it may work well with time-weighted since that's tilted towards recent changes, checking daily would catch them quickly, and bands would limit the rebalancing to when it is most needed.

This would introduce the variables of the exponential weight (rough equivalent of lookback period) and also the band size, and possibly be even more prone to over-fitting, but it would be interesting to see how sensitive it is to those things, and if it would have been any better than other proposed volatility based methods.

The only way I would consider this is if I could set up an automated check and alert system, and the bands didn't have to be so small to be effective that it would require rebalancing much more than monthly on average.