HEDGEFUNDIE's excellent adventure Part II: The next journey

[ocrtech]

I was curious how sensitive the cumulative results and drawdowns were to adjustments in the fixed AA allocation in OPs approach. I also wanted to compare that to the Inverse Volatility (Risk Parity) and Target Volatility approaches. The graphs below are based on the 1986-2019 simulated data supplemented with results through the end of July 2019. Each approach was run 20 times with daily data shifted by one day per run. Monthly rebalancing was performed for all approaches. Inverse Volatility is capped at the UPRO % allocation. Target Volatility is 20% target and UPRO is capped at UPRO % allocation.





In the backtest, OPs shift in allocation reduced his cumulative returns by about -6%(-75k) but increased his max drawdown from 46% to 61%.

[MotoTrojan]

I really don't get the reason why OP changed the strategy from risk parity 40/60 to 55/45 which seems random to me.

This is a comparison for the time period Jun 2003 to Nov 2009. I choose this time period because of the same SP500 P/E and the same 20-year rate as mentioned in viewtopic.php?f=10&t=272007&start=3300#p4690830. Also because it included a bear market. It seems to me that the 55/45 allocation performed worse. You can't only use the data from an equity bull market to predict the future...

What about 1955-1982, or 1955-2019.

[HEDGEFUNDIE]

I was curious how sensitive the cumulative results and drawdowns were to adjustments in the fixed AA allocation in OPs approach. I also wanted to compare that to the Inverse Volatility (Risk Parity) and Target Volatility approaches. The graphs below are based on the 1986-2019 simulated data supplemented with results through the end of July 2019. Each approach was run 20 times with daily data shifted by one day per run. Monthly rebalancing was performed for all approaches. Inverse Volatility is capped at the UPRO % allocation. Target Volatility is 20% target and UPRO is capped at UPRO % allocation.





In the backtest, OPs shift in allocation reduced his cumulative returns by about -6%(-75k) but increased his max drawdown from 46% to 61%.

How about starting the simulation at July 2012, the first time long rates were this low.

[physixfan]

I really don't get the reason why OP changed the strategy from risk parity 40/60 to 55/45 which seems random to me.

This is a comparison for the time period Jun 2003 to Nov 2009. I choose this time period because of the same SP500 P/E and the same 20-year rate as mentioned in viewtopic.php?f=10&t=272007&start=3300#p4690830. Also because it included a bear market. It seems to me that the 55/45 allocation performed worse. You can't only use the data from an equity bull market to predict the future...

What about 1955-1982, or 1955-2019.

In my opinion, we should not start with 1955. If I were in that time, I knew the inflation was high and there was no sign that the FED interest rate is going to be lower at that time, then there is no negative relation between stock and LTT, and I would not invest in this strategy. I would also quit this strategy once the inflation seems to be higher and FED can't control it any more.

[Lee_WSP]

Why do we keep on posting "hey, check out this tiny slice in time" graphs?

No one is planning on selling in X years. If you extend the slice in time one year forwards or backwards, the results are 99% of the time completely different.

Isn't a Monte Carlo simulation a better predictor of what the future may or may not hold?

[Kevin M]

Doing these calculations with same assumptions, except that yield drops from 3% at t0 to 2% at t1, we get:

ir = 3%
cr = 19.5%
tr = 22.5%

So much math, and yet missing the obvious, which is that we are using 3x leverage.

So just to be clear, there is no dispute about the correctness of the bond math then.

Remember the context of the discussion, which is that you can't get the same return going from 2% to 0% (or somewhat lower) that you get in going from 15% to 2%. That applies whether you're looking at 1X or 3X.

Backing up further, I'm simply providing the bond math for those who have cautioned against relying too much on the returns from a backtest period when long-term Treasuries (not leveraged, not even extended duration) returned 2,400%. I'm not saying that you can't make a lot of money on long-term Treasuries, and of course even more on a 3X LTT fund with a duration of about 54 years, if rates drop from 2% to 0%.

Kevin

[faf]

I do find it comical how your three backtest periods are during a bull market. How did 55/45 do overall...?

Over the time period 1955-2018, anyway, the returns were higher all the way up to 55% UPRO / 45% TMF. It's pretty bold, though. Little OGs might want to spring for the milder version at 50% UPRO / 50% TMF for a less bumpy ride in a bear market for stocks.



P1 is 55% UPRO.
P2 is 50% UPRO.
P3 is 45% UPRO.
P4 is 40% UPRO.
P5 is the S&P 500 index.

Thank you for this. How does 60% UPRO / 40% TMF fare in this comparison?

Also, any chance you can post those 6 for 1955-2010 as well, just to see what it might look like without the current massive bull market influencing gains in UPRO, thereby skewing the % allocation towards UPRO? Thanks!

[MoneyMarathon]

In my opinion, we should not start with 1955. If I were in that time, I knew the inflation was high and there was no sign that the FED interest rate is going to be lower at that time, then there is no negative relation between stock and LTT, and I would not invest in this strategy. I would also quit this strategy once the inflation seems to be higher and FED can't control it any more.

If you want to bet on the direction of yields, you can.

The 55% UPRO / 45% TMF choice reflects a more constrained view there - the idea that they're just not going to shoot upwards. It reduces exposure to the direction of yields and increases exposure to the direction of the stock market. This is consistent with the original aim of "supercharging" a bet on the US stock market. In the course of discussion, most people in this thread now know that the 60% TMF allocation contributed huge amounts to juice returns in the period 1982 or 1987 (or whatever - most of the starting points that people fixated on), to present.

The early discussion reflected this false belief that TMF was "just" there for rebalancing with negative correlation and wasn't a significant source of returns. The early discussion was extremely confused and misguided in that regard.

[Lee_WSP]

Doing these calculations with same assumptions, except that yield drops from 3% at t0 to 2% at t1, we get:

ir = 3%
cr = 19.5%
tr = 22.5%

So much math, and yet missing the obvious, which is that we are using 3x leverage.

So just to be clear, there is no dispute about the correctness of the bond math then.

Remember the context of the discussion, which is that you can't get the same return going from 2% to 0% (or somewhat lower) that you get in going from 15% to 2%. That applies whether you're looking at 1X or 3X.

Backing up further, I'm simply providing the bond math for those who have cautioned against relying too much on the returns from a backtest period when long-term Treasuries (not leveraged, not even extended duration) returned 2,400%. I'm not saying that you can't make a lot of money on long-term Treasuries, and of course even more on a 3X LTT fund with a duration of about 54 years, if rates drop from 2% to 0%.

Kevin

https://portfoliocharts.com/2019/05/27/ ... convexity/

[MotoTrojan]

Since everyone is so focused on such short timeframes now I thought I would chime in to say my XIRR since 2/27/19 is 88.9%, compared to an S&P500 price-return XIRR of 5.2% .

[Legitthe]

Doing these calculations with same assumptions, except that yield drops from 3% at t0 to 2% at t1, we get:

ir = 3%
cr = 19.5%
tr = 22.5%

So much math, and yet missing the obvious, which is that we are using 3x leverage.

So just to be clear, there is no dispute about the correctness of the bond math then.

Remember the context of the discussion, which is that you can't get the same return going from 2% to 0% (or somewhat lower) that you get in going from 15% to 2%. That applies whether you're looking at 1X or 3X.

Backing up further, I'm simply providing the bond math for those who have cautioned against relying too much on the returns from a backtest period when long-term Treasuries (not leveraged, not even extended duration) returned 2,400%. I'm not saying that you can't make a lot of money on long-term Treasuries, and of course even more on a 3X LTT fund with a duration of about 54 years, if rates drop from 2% to 0%.

Kevin

https://portfoliocharts.com/2019/05/27/ ... convexity/

Longer duration means more sensitivity to interest rates. Who knew

[MotoTrojan]

Doing these calculations with same assumptions, except that yield drops from 3% at t0 to 2% at t1, we get:

ir = 3%
cr = 19.5%
tr = 22.5%

So much math, and yet missing the obvious, which is that we are using 3x leverage.

So just to be clear, there is no dispute about the correctness of the bond math then.

Remember the context of the discussion, which is that you can't get the same return going from 2% to 0% (or somewhat lower) that you get in going from 15% to 2%. That applies whether you're looking at 1X or 3X.

Backing up further, I'm simply providing the bond math for those who have cautioned against relying too much on the returns from a backtest period when long-term Treasuries (not leveraged, not even extended duration) returned 2,400%. I'm not saying that you can't make a lot of money on long-term Treasuries, and of course even more on a 3X LTT fund with a duration of about 54 years, if rates drop from 2% to 0%.

Kevin

https://portfoliocharts.com/2019/05/27/ ... convexity/

Longer duration means more sensitivity to interest rates. Who knew

You are missing the point.

[rascott]

Since everyone is so focused on such short timeframes now I thought I would chime in to say my XIRR since 2/27/19 is 88.9%, compared to an S&P500 price-return XIRR of 5.2% .

In all the noise it's been lost.....but what are you actually doing? 3 different combos?

[Legitthe]

So much math, and yet missing the obvious, which is that we are using 3x leverage.

So just to be clear, there is no dispute about the correctness of the bond math then.

Remember the context of the discussion, which is that you can't get the same return going from 2% to 0% (or somewhat lower) that you get in going from 15% to 2%. That applies whether you're looking at 1X or 3X.

Backing up further, I'm simply providing the bond math for those who have cautioned against relying too much on the returns from a backtest period when long-term Treasuries (not leveraged, not even extended duration) returned 2,400%. I'm not saying that you can't make a lot of money on long-term Treasuries, and of course even more on a 3X LTT fund with a duration of about 54 years, if rates drop from 2% to 0%.

Kevin

https://portfoliocharts.com/2019/05/27/ ... convexity/

Longer duration means more sensitivity to interest rates. Who knew

You are missing the point.

I mean not really, even if I was being a bit of a smartass. We’re really just illustrating the negative relationship between interest rates and both duration and convexity.

[Lee_WSP]

I'm pretty sure the graph and the paper who made it shows that TSM can indeed keep on going even as rates go to zero and even negative. If rates start to rise, TMF is going to look pretty bad.

[NMBob]

Therefore, at these rates today, and with the three comparable historical periods I backtested above, I am hereby making the following change to the OG strategy:

55% UPRO / 45% TMF

This is not risk parity any more correct? I have a hard time believing that a dollar in 30 year treasuries is riskier than a dollar in the market.

It doesn't seem to be. edit; well it is for the time periods Although i wonder if there may be a trend across the industry to bastardize what it means.

Definition: Risk parity (or risk premia parity) is an approach to investment portfolio management which focuses on allocation of risk, usually defined as volatility, rather than allocation of capital. The risk parity approach asserts that when asset allocations are adjusted (leveraged or deleveraged) to the same risk level, the risk parity portfolio can achieve a higher Sharpe ratio and can be more resistant to market downturns than the traditional portfolio.

Some AA ratios with CAGRS were presented and a ratio was selected. If this was about risk parity, shouldn't numbers on volatility and sharp ratios be presented and a selection made from those? EDIT; after plugging in pv almost same dates for the only 3 time periods in , you get high sharps at/near his AA.

[ocrtech]

I was curious how sensitive the cumulative results and drawdowns were to adjustments in the fixed AA allocation in OPs approach. I also wanted to compare that to the Inverse Volatility (Risk Parity) and Target Volatility approaches. The graphs below are based on the 1986-2019 simulated data supplemented with results through the end of July 2019. Each approach was run 20 times with daily data shifted by one day per run. Monthly rebalancing was performed for all approaches. Inverse Volatility is capped at the UPRO % allocation. Target Volatility is 20% target and UPRO is capped at UPRO % allocation.





In the backtest, OPs shift in allocation reduced his cumulative returns by about -6%(-75k) but increased his max drawdown from 46% to 61%.

How about starting the simulation at July 2012, the first time long rates were this low.

Here you go. Same constraints aside from the time period which is now July 2012 through the end of July 2019. I guess the question is, how confident are you that this 7 year slice is indicative of what the next 20 years will be like. If you are wrong, the drawdowns are likely to be brutal.





This is what the drawdowns look like with all approaches set to a cap of 55% UPRO on a time period of May 1986 - July 2019.

[Legitthe]

Therefore, at these rates today, and with the three comparable historical periods I backtested above, I am hereby making the following change to the OG strategy:

55% UPRO / 45% TMF

This is not risk parity any more correct? I have a hard time believing that a dollar in 30 year treasuries is riskier than a dollar in the market.

It doesn't seem to be. Although i wonder if there may be a trend across the industry to bastardize what it means.

Risk parity (or risk premia parity) is an approach to investment portfolio management which focuses on allocation of risk, usually defined as volatility, rather than allocation of capital. The risk parity approach asserts that when asset allocations are adjusted (leveraged or deleveraged) to the same risk level, the risk parity portfolio can achieve a higher Sharpe ratio and can be more resistant to market downturns than the traditional portfolio.

Some AA ratios with CAGRS were presented and a ratio was selected. If this was about risk parity, shouldn't numbers on volatility and sharp ratios be presented and a selection made from those?

It can still be risk parity if OP believes that the ex post risk for TMF is understating the actual true risk that TMF is exposed to to. Doesn’t seem like too ridiculous assumption to make.

[NMBob]

well, it looks like he didn't just show those numbers. using pv but having to start at the begining of his start years, january as opposed to later month, for the 3 periods do seem to show data for std deviation and sharp that support his 3 time period risk parity AA for the specific periods. will edit in the pv data. std dev and sharp in bold.

jan2012-2019
Portfolio performance statistics
Portfolio Initial Balance Final Balance CAGR Stdev Best Year Worst Year Max. Drawdown Sharpe Ratio Sortino Ratio US Mkt Correlation
40/60u/t $10,000 $43,071 21.24% 20.77% 73.61% -16.67% -25.14% 1.01 1.67 0.40
50/50 $10,000 $52,654 24.49% 20.46% 67.67% -18.08% -22.64% 1.15 1.93 0.63
60u/40t $10,000 $63,205 27.52% 21.53% 61.74% -19.49% -25.18% 1.22 2.05 0.80

jan2015-2019
Portfolio Initial Balance Final Balance CAGR Stdev Best Year Worst Year Max. Drawdown Sharpe Ratio Sortino Ratio US Mkt Correlation
Portfolio 1 $10,000 $16,804 11.99% 22.26% 42.74% -16.67% -25.14% 0.58 0.89 0.43
Portfolio 2 $10,000 $18,167 13.91% 22.21% 47.05% -18.08% -22.64% 0.66 1.02 0.65
Portfolio 3 $10,000 $19,586 15.80% 23.55% 51.92% -19.49% -25.18% 0.71 1.09 0.81

jan2016-2019
Portfolio performance statistics
Portfolio Initial Balance Final Balance CAGR Stdev Best Year Worst Year Max. Drawdown Sharpe Ratio Sortino Ratio US Mkt Correlation
Portfolio 1 $10,000 $18,740 19.16% 20.67% 42.74% -16.67% -25.14% 0.90 1.36 0.49
Portfolio 2 $10,000 $20,071 21.46% 21.14% 47.05% -18.08% -21.40% 0.98 1.50 0.70
Portfolio 3 $10,000 $20,750 22.59% 21.86% 49.49% -18.79% -23.29% 1.00 1.54 0.78

[MotoTrojan]

Here you go. Same constraints aside from the time period which is now July 2012 through the end of July 2019. I guess the question is, how confident are you that this 7 year slice is indicative of what the next 20 years will be like. If you are wrong, the drawdowns are likely to be brutal.

An extra 10-15% drawdown in a major equity crisis isn't fun, but the original 40/60 has a much worse drawdown since 1955 due to the rising rate environment, so in that aspect the allocations shift actually reduces drawdown potential.

50/50 seems to be a sweet-spot based on your data. Given that your data was based on a period where long-bonds did exceptional, I don't think it is a big leap of faith to shift the allocation another 5% towards equity.

[MotoTrojan]

Since everyone is so focused on such short timeframes now I thought I would chime in to say my XIRR since 2/27/19 is 88.9%, compared to an S&P500 price-return XIRR of 5.2% .

In all the noise it's been lost.....but what are you actually doing? 3 different combos?

In all honesty I have been bouncing around much more than I'd have liked but the amount of data coming in has been significant.

I started with the OG 40/60 on 2/27/19 and had some additional contributions over the next couple of months.

I utilized the 20-day look-back risk-parity method for the month of July and then decided I wanted to reduce my exposure to long-bonds but maintain a passive rebalancing approach by doing 40/30/30 UPRO/TMF/EDV quarterly. On 8/1 I also split off 40% of that for a 20% target volatility UPRO/TMF combo which had even more economic research backing it than a look-back risk-parity.

The more I let it simmer though the more I realized the added strain of monthly rebalancing and having potential regret during a sudden downturn was not fitting in with my desire to simplify my finances overall. After seeing more mention of this 55/45 idea I did a deeper dive and found that I was comfortable with the additional risk, was getting the extra potential of return that I desired (even without volatility targeting), and liked the tilt away from long-bonds which was similar to my earlier 40/30/30 idea. The 45% TMF exposure is quite similar to the 30/30 TMF/EDV, so really I was just adding additional leverage to the equity side, at the cost of more volatility, larger drawdowns, and higher expected returns.

I do not plan to make any future changes but will be contributing some additional funds to PSLDX in the future (which would've gone into this) to diversify and keep overall portfolio leverage in check.

Yup, I am the poster-child of what not to do, but it has paid of handsomely and now I can buy & hold.

[rks]

Therefore, at these rates today, and with the three comparable historical periods I backtested above, I am hereby making the following change to the OG strategy:

55% UPRO / 45% TMF

Just curious what rates in future would prompt you to go back to 40% UPRO / 60% TMF? It seems 2 % long rate made you switch to the new AA, so trying to understand what are the rates where original AA makes sense. It seems anyone following this strategy would need to keep changing the AA as per the rates??

[dziuniek]

Wouldn't keeping UPRO at 40%, TMF at 50% and UGLD at 10% accomplish much of the same? Not sure about the upside, but it would curtail some of the risk present in TMF.

[NMBob]

'After seeing more mention of this 55/45 idea I did a deeper dive and found that I was comfortable with the additional risk,'

i am all confused. you argued this was risk parity, now state 'additional risk'

[MotoTrojan]

'After seeing more mention of this 55/45 idea I did a deeper dive and found that I was comfortable with the additional risk,'

i am all confused. you argued this was risk parity, now state 'additional risk'

Nowhere have I personally argued that 55/45 was risk-parity in the traditional or really any stance...?

[schismal]

Risk parity during 1987 - 2018 (LTT yield change from 7.4% to 3%)
UPRO: 40%
TMF: 60%

Risk parity during 1987 - 2007 (LTT yield change from 7.4% to 4.5%)
UPRO: 35%
TMF: 65%

Risk parity during 2008 - 2018 (LTT yield change from 4.5% to 3%)
UPRO: 50%
TMF: 50%

I think it's reasonable to argue that our simulated dataset may be giving a false indication of risk parity, given that LTT yields look fairly different today than from the majority of the backtesting on which that risk parity was calculated.

[HEDGEFUNDIE]

Risk parity during 1987 - 2018 (LTT yield change from 7.4% to 3%)
UPRO: 40%
TMF: 60%

Risk parity during 1987 - 2007 (LTT yield change from 7.4% to 4.5%)
UPRO: 35%
TMF: 65%

Risk parity during 2008 - 2018 (LTT yield change from 4.5% to 3%)
UPRO: 50%
TMF: 50%

I think it's reasonable to argue that our simulated dataset may be giving a false indication of risk parity, given that LTT yields look fairly different today than from the majority of the backtesting on which that risk parity was calculated.

Risk parity for 07/12 (the very first time long rates dropped into the 2% range) through 07/19 is 50/50.

[HEDGEFUNDIE]

Therefore, at these rates today, and with the three comparable historical periods I backtested above, I am hereby making the following change to the OG strategy:

55% UPRO / 45% TMF

Just curious what rates in future would prompt you to go back to 40% UPRO / 60% TMF? It seems 2 % long rate made you switch to the new AA, so trying to understand what are the rates where original AA makes sense. It seems anyone following this strategy would need to keep changing the AA as per the rates??

This is a great question, and I would answer it this way.

I don’t expect LTT rates to ever climb to the point where a 40/60 AA would make sense again. If rates do rise to that point, the strategy would be ruined anyway.

So I anticipate holding 55/45 for the foreseeable future.

[MoneyMarathon]

Wouldn't keeping UPRO at 40%, TMF at 50% and UGLD at 10% accomplish much of the same? Not sure about the upside, but it would curtail some of the risk present in TMF.

Yeah, anyone with a heavy position in long bonds should be open to taking a position in something else, like gold, that might do well if both bonds and stocks do not.

I think there's not much excitement about it overall - those who are more risk averse aren't 3x leveraged, and those that are less risk averse have any number of reasons not to like the added gold (conviction that significant rising yields are gone, low performer as an individual asset, lower performance starting with the very popular 1980s to present time period, trust in the idea that the original two ETF strategy has been "vetted," whatever that means) which might not be super great reasoning, but it is what it is.

3x leverage with only two already-volatile assets is incredibly risky. You're wise to consider adding a third and easing up, just a little, on the tail risk exposure. Others will just bet their conviction that this tail risk never shows up. Rates are apparently going to zero, negative infinity, and beyond (or at least staying well behaved and not going up much). So, no worries... if you believe that.

[Lee_WSP]

I hope that those of us employing this strategy are only risking a small portion of our assets. However, the other side of the coin is that even a 60/40 portfolio comes with risk.

There's also the risk of not having enough to retire. That's a risk whose medicine is to increase the savings rate, but what if we've got no more expenses to cut?

[Meaty]

asset allocation changed because risk parity now actually has little to do with the goal of this nice plan? (thanks so much for the work I have implemented a variant in a small allocation)

Changing AA which increases the probable loss in a bear market, as this does, seems to be the opposite of risk parity. Seems folks now looking at this as return/risk just like folks would pick a 60stock/40 bond AA instead of a risk parity 10stock/90 bond AA, which is fine with me.

In an equity bear market. Risk parity is not about limiting drawdowns but using historical precedence the 55/45 would have a much smaller drawdown than the original 40/60.

This seems counterintuitive- more exposure to UPRO would suggest lower returns in an equity bear market. Can you help me understand why this is not the case?

[MoneyMarathon]

However, the other side of the coin is that even a 60/40 portfolio comes with risk.

There's also the risk of not having enough to retire. That's a risk whose medicine is to increase the savings rate, but what if we've got no more expenses to cut?

One should take only so much risk (that may be a bit more than 60/40... but definitely less than 120/180 or 165/135 with 3x leverage...) with money you want to be there later for necessities.

If you're in that much of a shortfall, you gotta earn/save more or work longer or both. Or just cut expenses and boost benefits in retirement: retire later, at 70 if needed, and move to a lower COL area. Lots of people manage without saving or giving thought to the subject until late in life.

Not everybody gets to have $10 million and a yacht... or a certain amount of withdrawals per year or whatever the goal is.

I hope that those of us employing this strategy are only risking a small portion of our assets.

[Lee_WSP]

Risk of not having enough was just something that came to mind. Upon further reflection, those in that situation are unfortunately, the least suitable to employ this strategy.

[MotoTrojan]

asset allocation changed because risk parity now actually has little to do with the goal of this nice plan? (thanks so much for the work I have implemented a variant in a small allocation)

Changing AA which increases the probable loss in a bear market, as this does, seems to be the opposite of risk parity. Seems folks now looking at this as return/risk just like folks would pick a 60stock/40 bond AA instead of a risk parity 10stock/90 bond AA, which is fine with me.

In an equity bear market. Risk parity is not about limiting drawdowns but using historical precedence the 55/45 would have a much smaller drawdown than the original 40/60.

This seems counterintuitive- more exposure to UPRO would suggest lower returns in an equity bear market. Can you help me understand why this is not the case?

I’m saying in an equity bear more UPRO exposure is worse but that there can also be a bond bear and in that case more UPRO can help.

[HawkeyePierce]

I hope that those of us employing this strategy are only risking a small portion of our assets. However, the other side of the coin is that even a 60/40 portfolio comes with risk.

There's also the risk of not having enough to retire. That's a risk whose medicine is to increase the savings rate, but what if we've got no more expenses to cut?

I'm going in with a pretty small portion of my Roth assets (~$35k). It's equivalent to about 1/4 of my annual stock grant from work. Enough to be a meaningful sum a couple decades from now if this strategy pans out, but not so much to wreck me if it doesn't.

My savings rate is around 50% of my gross and I'm keeping the rest of my investments in a pretty standard Boglehead-ish 80/20 portfolio. Based on all of that, I'm confident that employing this strategy with such a sum does not pose a meaningful risk to my financial independence plans while also giving me a chance of reaching them far sooner.

[sfmurph]

I’ve been in the OG 60/40 since April. I was attracted by the elegance; the set-it-and-forget-it nature of the decision, the use of leverage, the quarterly rebalancing, the non-correlated assets, plus the approach to choosing the percentages. Initially, OP said "Over the long run, the average annual volatility of long term Treasuries has been 10%; unsurprisingly, the S&P 500 has been more volatile, at 15%. Therefore, a risk parity portfolio would hold more Treasuries than stocks. Simple arithmetic shows parity is achieved at 40/60."

I’ve been following some of the lookback conversations, and I see that Treasurie
s can’t move from 15% to 2% again, but the monthly lookback and rebalance didn’t
resonate with me.

I was surprised to read HEDGEFUNDIE’s move to 45/55. The arithmetic isn’t so sim
ple anymore. By changing the calculation period, what does the volatility become
? It’s not 10% / 15%, it seems. But what is it? And what’s the period that’s bee
n blessed? 1995 to 1980? Something else? It’s lost its elegance.

Another concern with the 45/55 is with MoneyMarathon’s 45/55 CASHZERO/UPRO simulation that shows that 55% UPRO with 45% flat is that you get the returns of the US stock market amplified by another position (TMF). But if the hesitation is that TMF will provide a negative return, why not get the US stock market return in a simpler fashion?

The OP updated the starting post to point to a message he wrote in March. I get that it’s been a process, but the modifaction of vision hasn’t been expressed as well as the initial concept.

Now, my quarterly rebalance date is coming up on Friday, so please help!

[Nickel & Dime]

OP updated with the AA change. My rebalance went through this morning after yet another LTT rate drop. Let's hope I sold at the top!

I have a different thing going on with EDV and UPRO, but anyway I changed my allocations yesterday (Sun Aug 11), and today my rebalance didn’t happen. It has the change saved which shows the new target allocation, but still has the actual old allocation. Quite concerning, indeed. I sent them (M1) a message today.

[MotoTrojan]

OP updated with the AA change. My rebalance went through this morning after yet another LTT rate drop. Let's hope I sold at the top!

I have a different thing going on with EDV and UPRO, but anyway I changed my allocations yesterday (Sun Aug 11), and today my rebalance didn’t happen. It has the change saved which shows the new target allocation, but still has the actual old allocation. Quite concerning, indeed. I sent them (M1) a message today.

If you don’t hit rebalance or add new funds it won’t do anything.

[HEDGEFUNDIE]

OP updated with the AA change. My rebalance went through this morning after yet another LTT rate drop. Let's hope I sold at the top!

I have a different thing going on with EDV and UPRO, but anyway I changed my allocations yesterday (Sun Aug 11), and today my rebalance didn’t happen. It has the change saved which shows the new target allocation, but still has the actual old allocation. Quite concerning, indeed. I sent them (M1) a message today.

You have to actually hit the Rebalance button, not just change the allocations.

[Nickel & Dime]

OP updated with the AA change. My rebalance went through this morning after yet another LTT rate drop. Let's hope I sold at the top!

I have a different thing going on with EDV and UPRO, but anyway I changed my allocations yesterday (Sun Aug 11), and today my rebalance didn’t happen. It has the change saved which shows the new target allocation, but still has the actual old allocation. Quite concerning, indeed. I sent them (M1) a message today.

You have to actually hit the Rebalance button, not just change the allocations.

Ok, thanks! I didn’t do that....

[HEDGEFUNDIE]

The OP updated the starting post to point to a message he wrote in March.

Fixed.

[caklim00]

Oh my... I go on vacation, miss half a week and I see this thread has gone nuts. Going to stick with the 1/3 40/60, 1/3 20 day look back for risk parity, and 1/3 20% target volatility for UPRO. Too much thinking going on here.

[drock]

If we're worried about bond rates and duration and accepting that we're not going to make as much on your bond positions going forward, would revisiting using TYD over TMF have any merit?

[MoneyMarathon]

If we're worried about bond rates and duration and accepting that we're not going to make as much on your bond positions going forward, would revisiting using TYD over TMF have any merit?

EDV (extended duration treasuries) tends to outperform TYD (3x intermediate) slightly when yields are flat or increasing, apparently due to the Cost Matters Hypothesis (lower expense ratio). TYD might have a small edge when yields are falling (but not much). EDV has a much smaller bid-ask spread (lower trading costs) and lower tracking error... which seems to tip the scales definitively over to EDV (any time a Vanguard ETF can play a similar role as a Direxion ETF, prefer Vanguard IMO).

[privatefarmer]

Holy cow. HEDGEFUNDIE started this thread which has probably been the busiest, most popular thread of all time on bogleheads, has spread to other sites (reddit) and even now been mentioned on podcasts like Animal Spirits which is ran by the CIO of a major wealth management firm (Ritholtz wealth mgmt).... now the guy wants to SLIGHTLY adjust his allocation after much research and evidence has been pooring in and people are criticizing him for it??

I’ve chamged my allocation several times since the beginning of this thread and, news flash, I’m doing this with 100% of my portfolio. I believe in buy and hold but I also believe that as more evidence comes in you’d be foolish not to adjust. Thanks to HEDGEFUNDIE, I’ve discovered risk parity, read several books on it now, and am fully bought in. I currently am employing the 20day risk parity look back strategy with a targeted 4.5x total leverage using a combo of LETFs and margin loan from IB. I also consider my house, which is a considerable part of my net worth, as a substitute for gold (not perfect but hopefully would keep pace with inflation).

Anyhow, give the guy a break. HEDGEFUNDIE, I’ve made out like a fat cat since you started this thread so I for one would like to THANK YOU and god speed. I made $23,000 just TODAY so thank you. I am holding the bull by the horns let’s hope it’s a fun ride.

[HEDGEFUNDIE]

Holy cow. HEDGEFUNDIE started this thread which has probably been the busiest, most popular thread of all time on bogleheads, has spread to other sites (reddit) and even now been mentioned on podcasts like Animal Spirits which is ran by the CIO of a major wealth management firm (Ritholtz wealth mgmt).... now the guy wants to SLIGHTLY adjust his allocation after much research and evidence has been pooring in and people are criticizing him for it??

I’ve chamged my allocation several times since the beginning of this thread and, news flash, I’m doing this with 100% of my portfolio. I believe in buy and hold but I also believe that as more evidence comes in you’d be foolish not to adjust. Thanks to HEDGEFUNDIE, I’ve discovered risk parity, read several books on it now, and am fully bought in. I currently am employing the 20day risk parity look back strategy with a targeted 4.5x total leverage using a combo of LETFs and margin loan from IB. I also consider my house, which is a considerable part of my net worth, as a substitute for gold (not perfect but hopefully would keep pace with inflation).

Anyhow, give the guy a break. HEDGEFUNDIE, I’ve made out like a fat cat since you started this thread so I for one would like to THANK YOU and god speed. I made $23,000 just TODAY so thank you. I am holding the bull by the horns let’s hope it’s a fun ride.

Just to be clear, I do not endorse this kind of behavior.

But you’re welcome!

[privatefarmer]

Holy cow. HEDGEFUNDIE started this thread which has probably been the busiest, most popular thread of all time on bogleheads, has spread to other sites (reddit) and even now been mentioned on podcasts like Animal Spirits which is ran by the CIO of a major wealth management firm (Ritholtz wealth mgmt).... now the guy wants to SLIGHTLY adjust his allocation after much research and evidence has been pooring in and people are criticizing him for it??

I’ve chamged my allocation several times since the beginning of this thread and, news flash, I’m doing this with 100% of my portfolio. I believe in buy and hold but I also believe that as more evidence comes in you’d be foolish not to adjust. Thanks to HEDGEFUNDIE, I’ve discovered risk parity, read several books on it now, and am fully bought in. I currently am employing the 20day risk parity look back strategy with a targeted 4.5x total leverage using a combo of LETFs and margin loan from IB. I also consider my house, which is a considerable part of my net worth, as a substitute for gold (not perfect but hopefully would keep pace with inflation).

Anyhow, give the guy a break. HEDGEFUNDIE, I’ve made out like a fat cat since you started this thread so I for one would like to THANK YOU and god speed. I made $23,000 just TODAY so thank you. I am holding the bull by the horns let’s hope it’s a fun ride.

Just to be clear, I do not endorse this kind of behavior.

But you’re welcome!

LOL. I’ll either name my yacht after you or my 1987 Dodge Dart, depending on how things go.

[sfmurph]

The OP updated the starting post to point to a message he wrote in March.

Fixed.

Ah, OK. That makes more sense. However, you quote CAGR in these 3 examples, and nothing about volatility or risk parity. I don't see anything in these numbers where 45/55 has a logic to it. Maybe it seems like over-fitting?

And yes, thank you. This is an incredibly interesting approach. It's really gotten me to appreciate the ballast of bonds and to understand what I'm comfortable with with them. Thank you!

[MoneyMarathon]

4.5x total leverage using a combo of LETFs and margin loan from IB

Might be time to take some winnings off the table (i.e., cut down on margin)? If this says what it looks like it says, your ETFs are at IB and they can start selling your assets at any time if your equity in the account (the part you own - beyond the money loaned to you by IB) doesn't meet their minimum requirements. Even under relatively favorable historical conditions for a bear market, like 2000-2003 and 2008, there were some drawdowns for a UPRO/TMF combo that could trigger margin calls.

What's the big hurry? Beyond a certain point of leverage and with a certain amount of time in the market, a more stable and less-levered portfolio will always win (not just deliver smaller swings - deliver more performance), with what most people could just go ahead and call total and complete certainty... you need to break out several decimal places to get the odds of it not happening.

[MotoTrojan]

If we're worried about bond rates and duration and accepting that we're not going to make as much on your bond positions going forward, would revisiting using TYD over TMF have any merit?

EDV (extended duration treasuries) tends to outperform TYD (3x intermediate) slightly when yields are flat or increasing, apparently due to the Cost Matters Hypothesis (lower expense ratio). TYD might have a small edge when yields are falling (but not much). EDV has a much smaller bid-ask spread (lower trading costs) and lower tracking error... which seems to tip the scales definitively over to EDV (any time a Vanguard ETF can play a similar role as a Direxion ETF, prefer Vanguard IMO).

Correct. Furthermore, my backtesting suggests that there is no significant difference between how you get your weighted-duration exposure, so the real question is what weighted-duration you want. Using TYD at the same allocation as you would've used TMF will simply reduce your duration exposure; yes it will cause less of a drag/loss during rising rate periods, but it also will have less protection during an equity crisis, and less rebalancing bonus when correlations are negative; not that it is a bad move, just a bad way to implement it. There may be some merit to going ultra-short with mega leverage (like the individuals using 2-year treasury futures) but my backtesting only looked at ETFs that may be used in a manner similar to the OG strategy.

As noted, there really is (in my humble opinion) no use for TYD since you can replicate any use of it with a combo of TMF & a lower duration fund such as IEF, TLT, and/or EDV, reducing your overall exposure to borrowing costs and high expense ratios. By using TYD you are simply increasing the amount of high-cost leveraged funds you need to hold to get your desired duration exposure.

In general EDV is a very interesting fund as it gets pretty good duration exposure (about 1/2 the effectiveness as TMF) with no leverage costs or high expenses. ~40/60 UPRO/EDV is a pretty balanced risk profile, but I am greedy .