HEDGEFUNDIE's excellent adventure Part II: The next journey

[RandomWord]

I started this adventure in March 2019 with the original allocation of 40/60. Switched allocation to UPRO 55/ TMF 45 when HEDGEFUNDIE made that recommendation. I have rebalanced every quarter. Well as of today I have doubled my investment. I just wanted to thank HEDGEFUNDIE and all others that have contributed to this excellent adventure.Thanks again!

Congratulations! I am only up 58% myself. Chalk it up to poor rebalancing timing

At 58% in 1 year, do you kinda giggle under your breath at all the conversation about tweaking the strategy? Like, WHAT ELSE DO YOU GUYS WANT?? Or are you more proud that you started some [offensive language removed by admin LadyGeek] worth that much conversation?

you're not at all scared by the massive crash it took in march, or that almost all the gains have come entirely from long-term bonds, which are now near zero yield?

Real return of VUSTX from 1987-2003 : 5.4% CAGR
Real return of VUSTX from 2004-today : 5.2% CAGR

in 1987 the 20 and 30 year yield were more than a smidge higher than they are today. Starting yield is irrelevant. I repeat, starting yield is irrelevant. A bond FUND gets its return from multiple sources, capital appreciation due to a steep yield curve being one of them.

I realize bond yields can go negative, but it seems crazy that you expect them to go keep going further and further negative like it doesn't matter. Who's going to pay 10% interest for the privilege of loaning out their money?

[jaj2276]

Rebalance day for me today. 1 slice went to 14/86, the other went to 30/70, and the OG remains at 40/60 (has drifted to 46/54 and was not rebalanced). Portfolio is up 71% since Feb 15 '19.

[PeanutButter]

I started this adventure in March 2019 with the original allocation of 40/60. Switched allocation to UPRO 55/ TMF 45 when HEDGEFUNDIE made that recommendation. I have rebalanced every quarter. Well as of today I have doubled my investment. I just wanted to thank HEDGEFUNDIE and all others that have contributed to this excellent adventure.Thanks again!

Congratulations! I am only up 58% myself. Chalk it up to poor rebalancing timing

What is the optimal rebalance plan?

Once per quarter or other period of time?

Once you drift more than x %?

One thought I had is a more frequent rebalance when things are crazy, perhaps something based on VIX (this might accomplish some of what was recently discussed on adding VIX to the strategy).

[Walkure]

I realize bond yields can go negative, but it seems crazy that you expect them to go keep going further and further negative like it doesn't matter. Who's going to pay 10% interest for the privilege of loaning out their money?

I think most insurance companies are effectively required to keep some portion of their float in risk-free* assets. Maybe we'll see insurance premiums start rising to compensate for the losses on float. What a time to be alive...

*Of course, at negative rates, risk-free simply means "guaranteed to lose nominal dollars."

[LittleBitMore]

I realize bond yields can go negative, but it seems crazy that you expect them to go keep going further and further negative like it doesn't matter. Who's going to pay 10% interest for the privilege of loaning out their money?

I think most insurance companies are effectively required to keep some portion of their float in risk-free* assets. Maybe we'll see insurance premiums start rising to compensate for the losses on float. What a time to be alive...

*Of course, at negative rates, risk-free simply means "guaranteed to lose nominal dollars."

As someone who works in insurance, this is correct -- bond (particularly government bond) investments will always be a very core (i.e. large) portion of the portfolio (irrespective of rate/yield) because of regulation. And yes, rates will go up because of it -- you make returns through underwriting & investment, when one goes down the other must go up or you underperform/lose money.

[hilink73]

End of April rebalancing done yesterday for my target volatility 25% strategy using PV.

Feb UPRO 42% / TMF 58% Perf.: -9.3%
Mar UPRO 9% / TMF 91% Perf.: -10.9%
Apr UPRO 21% / TMF 79% Perf.: 6.7%


Let's see what May brings.

[occambogle]

What rebalancing strategy is generally recommended with the standard 55/45 UPRO/TMF? My understanding was that quarterly was most generally recommended... but should it be more frequent in times of high volatility?

[Forester]

How did the default strategy perform from a terrible recent starting point - January 2000?

[neo_ny]

FDFAX and FSUTX go back to 1985

What levels were UPRO/TMF at when you started the adventure? Did their levels have any bearing on when to start it?

[KSActuary]

I realize bond yields can go negative, but it seems crazy that you expect them to go keep going further and further negative like it doesn't matter. Who's going to pay 10% interest for the privilege of loaning out their money?

I think most insurance companies are effectively required to keep some portion of their float in risk-free* assets. Maybe we'll see insurance premiums start rising to compensate for the losses on float. What a time to be alive...

*Of course, at negative rates, risk-free simply means "guaranteed to lose nominal dollars."

As someone who works in insurance, this is correct -- bond (particularly government bond) investments will always be a very core (i.e. large) portion of the portfolio (irrespective of rate/yield) because of regulation. And yes, rates will go up because of it -- you make returns through underwriting & investment, when one goes down the other must go up or you underperform/lose money.

Large blocks of legacy insurance policies have fairly high minimum internal guaranteed rates to be credited (3%). I remember when the 30 year went under 3% the first time, big gasp from the life insurers as they need to match assets and liabilities. No profit for you.

[Forester]

Just started the Hedgefundie Adventure three weeks ago with my son’s account (Adaptive Allocation, Min Variance allocation) up 15% so far, D A A A A N G, wishing my entire IRA was in the past three weeks.

Allocating more to the strategy on 5/01/2020, but with QLD/UBT.

* thanks for the messaged link, I'm not certain my PMs are being sent.

[TheoLeo]

In the original thread I saw the backtests on page 22 showing the really bad performance from 1955 to 1983. Was this because of the oil crises and the inflation it caused?

[HoldTheLine]

For all those worried about the excellent adventure's exposure to LTTs, what if I told you there was another 2-asset strategy that:

• Performed similarly to UPRO/TMF in the past decade -- except with no negative years!
  
• Has been around for a decade longer - more data to backtest!
• Much lower expense ratio!
• Had a 22% CAGR in the 2000s!
• Can be rebalanced just once a year with no significant drop in CAGR. Set and forget!
• Has 0 reliance on LTTs -- or any kind of bonds for that matter!

What's this magic 2-asset portfolio? Here it is: Magic Portfolio! Never lose to Mr. Market again

In all seriousness, I thought this might be a useful comparison, especially if you're considering starting an Excellent Adventure solely based on the backtests.

[manlymatt83]

I’ve been doing 55/45 for about 6 months. In reading some of the recent posts on here, would love to do adaptive allocation once per month. Is there a spreadsheet or something to easily calculate this? Or does it require a bunch of manual steps via portfolio viz?

[randyharris]

I’ve been doing 55/45 for about 6 months. In reading some of the recent posts on here, would love to do adaptive allocation once per month. Is there a spreadsheet or something to easily calculate this? Or does it require a bunch of manual steps via portfolio viz?

Excel, here’s the meat and potatoes how to calculate:

https://www.bogleheads.org/forum/viewt ... 0#p5211022

[physixfan]

I’ve been doing 55/45 for about 6 months. In reading some of the recent posts on here, would love to do adaptive allocation once per month. Is there a spreadsheet or something to easily calculate this? Or does it require a bunch of manual steps via portfolio viz?

I wrote some python code to calculate the allocation after Portfolio Visualizer is not free any more.

[Noobvestor]

I tried reading the first few posts of the original thread and skimming this sequel, but 92 pages is a lot (and that's just Part II!). Can someone TL;DR this for me? Seems like UPRO has imploded. TMF has done pretty well during the implosion. Why so much interest in this strategy though?!

[Meaty]

I tried reading the first few posts of the original thread and skimming this sequel, but 92 pages is a lot (and that's just Part II!). Can someone TL;DR this for me? Seems like UPRO has imploded. TMF has done pretty well during the implosion. Why so much interest in this strategy though?!

Why so much interest? Probably ~20% CAGR with a better sharpe ratio and lower draw downs then holding the S&P

[Noobvestor]

I tried reading the first few posts of the original thread and skimming this sequel, but 92 pages is a lot (and that's just Part II!). Can someone TL;DR this for me? Seems like UPRO has imploded. TMF has done pretty well during the implosion. Why so much interest in this strategy though?!

Why so much interest? Probably ~20% CAGR with a better sharpe ratio and lower draw downs then holding the S&P

Eyeballs chart ... I'm seeing about a 75% peak-to-trough for UPRO this year. I get that is mitigated somewhat by eyeballs chart a roughly 50% gain in TMF, but ... OK. Not really blowing my mind here, but maybe have to read upthread for the bigger picture. Right now it looks like a combination of two asset classes that have done well this decade, each leveraged to the hilt, but I confess I've just glanced so far. I guess what I'm curious about is why one should expect two of the best-performing 2010s asset classes (US large/LTTs) to keep outperforming going forward

I guess I get the idea of super-leveraging to turbo-charge a portfolio, but the specific selections seem biased toward recency. Now US large has high valuations compared to small, value, developed, emerging, etc... and long bond rates are at/near record all-time lows. Where to next?!

[firebirdparts]

I tried reading the first few posts of the original thread and skimming this sequel, but 92 pages is a lot (and that's just Part II!). Can someone TL;DR this for me? Seems like UPRO has imploded. TMF has done pretty well during the implosion. Why so much interest in this strategy though?!

You're just spitballing and expecting us to respond. There's no mystery. If somebody loses a bunch of money they'll probably say something, but nobody has.

[nisiprius]

...I guess what I'm curious about is why one should expect two of the best-performing 2010s asset classes (US large/LTTs) to keep outperforming going forward ...

HEDGEFUNDIE has answered this question in ways that are satisfying to him, though unconvincing to me. His very first posting in his first thread, HEDGEFUNDIE's excellent adventure [risk parity strategy using 3X leveraged ETFs, is a summary of his thinking. The correlation between stocks and long-term Treasuries historically over all available data has been essentially zero, but if you look at the correlation between VTI and TLT on PortfolioVisualizer, it shows as -0.48 because it is only covering 2009-present. HEDGEFUNDIE's interpretation is that the correlation between stocks and bonds changed from positive to negative at some point, but that although it changed before, the change is permanent. He ties this to what he believes are permanent changes in the Fed's management policies, e.g. Volcker.

If you accept the premise that stocks and long-term Treasury bonds "have" negative correlation--not just zero, not just independent, but negative; not just "have had," but "have;" reliably moving in opposite directions, all the time, good times and bad, past and future--then it follows logically that long-term Treasury actually erase risk in stocks while having positive returns of their own. This, to me, is the financial equivalent of perpetual motion, but it is logical if you accept the premises. If you do, then it also follows that you ought to leverage up the bonds because they don't have enough covariance to erase all the risk. It also follows that you are doing the MPT-miracle thing of reducing the risk of the whole portfolio by increasing the risk in part of it. And, finally, if you have succeeded in reducing the risk without a commensurate reduction in return, now that you've reduced the risk you can take the whole thing and leverage it up without exceeding your risk tolerance.

[Noobvestor]

I tried reading the first few posts of the original thread and skimming this sequel, but 92 pages is a lot (and that's just Part II!). Can someone TL;DR this for me? Seems like UPRO has imploded. TMF has done pretty well during the implosion. Why so much interest in this strategy though?!

You're just spitballing and expecting us to respond. There's no mystery.

I was hoping for a broad-strokes take from those who have followed it on the thread's popularity. Maybe that's an unfair ask, and it's up to me to buckle down and read 100+ pages across two threads. If so, that's fine. I'll bookmark it for later. I just keep seeing it surface again and again, and have tried to get into it, but am a bit stumped by its appeal. Not trying to suggest any mystery, just seeing if I can grasp what's going on here.

If somebody loses a bunch of money they'll probably say something, but nobody has.

Well, this is part of my confusion. If someone like OP is relatively evenly split, give or take a bit, between (again, I'm rounding/eyeballing) Asset Class A that is down 75% and Asset Class B that is up 50%, haven't they in fact lost a lot recently? Tone is hard to tell online, but you seem annoyed - if the ask is annoying, please accept my apologies. Just trying to wrap my brain around this in general terms. No annoyance intended.

[nisiprius]

I admire HEDGEFUNDIE for posting the results of his strategy through March 18th in the first post of this thread, here, and hope he will continue to keep it updated. It's not trivially easy to follow it independently because PortfolioVisualizer only has monthly data and because I'm not completely sure I understand how it is to be rebalanced.

[Noobvestor]

...I guess what I'm curious about is why one should expect two of the best-performing 2010s asset classes (US large/LTTs) to keep outperforming going forward ...

HEDGEFUNDIE has answered this question in ways that are satisfying to him, though unconvincing to me. His very first posting in his first thread, HEDGEFUNDIE's excellent adventure [risk parity strategy using 3X leveraged ETFs, is a summary of his thinking. The correlation between stocks and long-term Treasuries historically over all available data has been essentially zero, but if you look at the correlation between VTI and TLT on PortfolioVisualizer, it shows as -0.48 because it is only covering 2009-present. HEDGEFUNDIE's interpretation is that the correlation between stocks and bonds changed from positive to negative at some point, but that although it changed before, the change is permanent. He ties this to what he believes are permanent changes in the Fed's management policies, e.g. Volcker.

If you accept the premise that stocks and long-term Treasury bonds "have" negative correlation--not just zero, not just independent, but negative; not just "have had," but "have;" reliably moving in opposite directions, all the time, good times and bad, past and future--then it follows logically that long-term Treasury actually erase risk in stocks while having positive returns of their own. This, to me, is the financial equivalent of perpetual motion, but it is logical if you accept the premises. If you do, then it also follows that you ought to leverage up the bonds because they don't have enough covariance to erase all the risk. It also follows that you are doing the MPT-miracle thing of reducing the risk of the whole portfolio by increasing the risk in part of it. And, finally, if you have succeeded in reducing the risk without a commensurate reduction in return, now that you've reduced the risk you can take the whole thing and leverage it up without exceeding your risk tolerance.

Thanks, Nis, this is exactly what I was looking for - really appreciate you taking the time to distill it down. I just keep seeing this thread surface, and then saw a the recent chart, which threw me and simultaneously got me intrigued about it all. My gut reaction is aligned with yours - negative correlations are fickle, and may not be here to stay. If they are, perhaps there's a free lunch in all of this, but history suggests to me that's unlikely. Anyway, thanks again for boiling it down. With this framework I feel like I can start skimming with a more knowledgeable baseline!

[perfectuncertainty]

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!

How often do you rebalance?

[KSActuary]

...I guess what I'm curious about is why one should expect two of the best-performing 2010s asset classes (US large/LTTs) to keep outperforming going forward ...

HEDGEFUNDIE has answered this question in ways that are satisfying to him, though unconvincing to me. His very first posting in his first thread, HEDGEFUNDIE's excellent adventure [risk parity strategy using 3X leveraged ETFs, is a summary of his thinking. The correlation between stocks and long-term Treasuries historically over all available data has been essentially zero, but if you look at the correlation between VTI and TLT on PortfolioVisualizer, it shows as -0.48 because it is only covering 2009-present. HEDGEFUNDIE's interpretation is that the correlation between stocks and bonds changed from positive to negative at some point, but that although it changed before, the change is permanent. He ties this to what he believes are permanent changes in the Fed's management policies, e.g. Volcker.

If you accept the premise that stocks and long-term Treasury bonds "have" negative correlation--not just zero, not just independent, but negative; not just "have had," but "have;" reliably moving in opposite directions, all the time, good times and bad, past and future--then it follows logically that long-term Treasury actually erase risk in stocks while having positive returns of their own. This, to me, is the financial equivalent of perpetual motion, but it is logical if you accept the premises. If you do, then it also follows that you ought to leverage up the bonds because they don't have enough covariance to erase all the risk. It also follows that you are doing the MPT-miracle thing of reducing the risk of the whole portfolio by increasing the risk in part of it. And, finally, if you have succeeded in reducing the risk without a commensurate reduction in return, now that you've reduced the risk you can take the whole thing and leverage it up without exceeding your risk tolerance.

Thanks, Nis, this is exactly what I was looking for - really appreciate you taking the time to distill it down. I just keep seeing this thread surface, and then saw a the recent chart, which threw me and simultaneously got me intrigued about it all. My gut reaction is aligned with yours - negative correlations are fickle, and may not be here to stay. If they are, perhaps there's a free lunch in all of this, but history suggests to me that's unlikely. Anyway, thanks again for boiling it down. With this framework I feel like I can start skimming with a more knowledgeable baseline!

Your signature most likely predicts that you will not like this strategy.

[Chicken Little]

nm...

[manlymatt83]

I’ve been doing 55/45 for about 6 months. In reading some of the recent posts on here, would love to do adaptive allocation once per month. Is there a spreadsheet or something to easily calculate this? Or does it require a bunch of manual steps via portfolio viz?

I wrote some python code to calculate the allocation after Portfolio Visualizer is not free any more.

Is that open source?

[TNWoods]

I admire HEDGEFUNDIE for posting the results of his strategy through March 18th in the first post of this thread, here, and hope he will continue to keep it updated. It's not trivially easy to follow it independently because PortfolioVisualizer only has monthly data and because I'm not completely sure I understand how it is to be rebalanced.

+1

If you Adventurers want to keep your ratings up, we need results. March 18th is like a light year ago?

This was a huge hit early, but has turned into a real snoozer.

If I recall correctly, HEDGEFUNDIE promised to update every time he clicked the re-balance button, and that means once per quarter. I believe he has done that. The next rebalance should be around June 18th. Also, a light year is a measurement of distance, not time.

TNWoods

[DosCommas]

I am following the 55/45 strategy. I started 40/60 with 100k which was invested over 3 quarters starting March 2019 (due to me wanting to avoid ST gains in taxable during rebalances the first year). I let the portfolio drift up a bit to 50/50 following OP’s recommendation to switch to 55/45. I rebalanced on Feb 24 to 55/45. While the timing wasn’t perfect my account sits now at $149k (33% UPRO/SPXL, 66% TMF). I have tax loss harvested, net about a $20k loss so far by swapping between UPRO and SPXL from time to time which will save me some taxes when I have to rebalance next month. So far this has way outperformed my 50/50 US/Intl 80/20 3-fund portfolio which has 10x my assets. I’m never going to add more money into the excellent adventure but it does seem possible right now that it outperforms the rest of my portfolio. It’s been a good hedge in this downturn as LTT has done well and S&P 500 held up better than global equities overall.

[HEDGEFUNDIE]

...I guess what I'm curious about is why one should expect two of the best-performing 2010s asset classes (US large/LTTs) to keep outperforming going forward ...

HEDGEFUNDIE has answered this question in ways that are satisfying to him, though unconvincing to me. His very first posting in his first thread, HEDGEFUNDIE's excellent adventure [risk parity strategy using 3X leveraged ETFs, is a summary of his thinking. The correlation between stocks and long-term Treasuries historically over all available data has been essentially zero, but if you look at the correlation between VTI and TLT on PortfolioVisualizer, it shows as -0.48 because it is only covering 2009-present. HEDGEFUNDIE's interpretation is that the correlation between stocks and bonds changed from positive to negative at some point, but that although it changed before, the change is permanent. He ties this to what he believes are permanent changes in the Fed's management policies, e.g. Volcker.

If you accept the premise that stocks and long-term Treasury bonds "have" negative correlation--not just zero, not just independent, but negative; not just "have had," but "have;" reliably moving in opposite directions, all the time, good times and bad, past and future--then it follows logically that long-term Treasury actually erase risk in stocks while having positive returns of their own. This, to me, is the financial equivalent of perpetual motion, but it is logical if you accept the premises. If you do, then it also follows that you ought to leverage up the bonds because they don't have enough covariance to erase all the risk. It also follows that you are doing the MPT-miracle thing of reducing the risk of the whole portfolio by increasing the risk in part of it. And, finally, if you have succeeded in reducing the risk without a commensurate reduction in return, now that you've reduced the risk you can take the whole thing and leverage it up without exceeding your risk tolerance.

This synthesis doesn't get my thesis quite right.

1. I am willing to grant that LTTs and stocks have a long run correlation of zero - EXCEPT in instances where stocks crash, in which case LTTs will increase. And yes, I am positing that this flight-to-safety effect will persist, and is a foundational element of the strategy.

2. The other foundational element of the strategy is that the Fed permanently changed monetary policy post-1982. Hence, LTTs will never go through another "cycle" of interest rate rise back to 1970s levels. This element is unrelated to the zero / negative correlation element. If we can accept this, then LTTs' role in the strategy is to purely serve as a stock crash and deflation insurance policy, with little risk of dragging down returns on its own.

[Chicken Little]

Big fan.

For the slow-learners;

1. Does the capacity of TMF to rise during a stock crash change as yields for LTT drop? I'm not sure how to phrase it better. Does the stock-crash-buffering-capacity of TMF decline as yields move closer to zero?

2. What would be the effect of yield curve control if that's ever adopted?

Thanks

[TNWoods]

Well, this is part of my confusion. If someone like OP is relatively evenly split, give or take a bit, between (again, I'm rounding/eyeballing) Asset Class A that is down 75% and Asset Class B that is up 50%, haven't they in fact lost a lot recently? Tone is hard to tell online, but you seem annoyed - if the ask is annoying, please accept my apologies. Just trying to wrap my brain around this in general terms. No annoyance intended.

Apply that logic to the true boglehead portfolio. Haven't all the true-blue bogleheads lost a lot of portfolio value recently?

Yes.

But the reason that doesn't matter to either group is that one of the tenets is that both strategies are long-term. Focusing on, or making decisions/judgements about the strategy based on short-term observations is not valuable or relevant.

TNWoods

[privatefarmer]

HEDGEFUNDIE, I know you have elaborated before but cannot find it in these 2,000,000 pages of thread... could you remind us what happened post 1982 as far as fed monetary policy and how that would prevent a 1970-ish spike in rates? Or any articles on the subject? Thanks!

[cos]

HEDGEFUNDIE, I know you have elaborated before but cannot find it in these 2,000,000 pages of thread... could you remind us what happened post 1982 as far as fed monetary policy and how that would prevent a 1970-ish spike in rates? Or any articles on the subject? Thanks!

Not OP, but from the Federal Reserve of San Francisco: https://www.frbsf.org/education/publica ... 70s-1980s/

And from the Board of the Federal Reserve System: https://www.federalreserve.gov/monetary ... policy.htm

[Lordosis]

Everybody seems to be freaking because interest rates are in the toilet. Correct me if I'm wrong but are they any lower then during the GFC? This portfolio did very well the decade following that. What am I missing?

[vijaym73]

10 year treasury was as low as 2.25% during GFC

30 year treasury low was 2.69%

I hate to say it but i feel negative rates will be coming.

VJ

[evoila]

Maybe this has been posted, but there are a lot of pages and despite the shelter-in-place, I really have better things to do. Is there a way to back test to determine what are the historically best days/months/etc. to contribute & re-balance a portfolio? Would be curious what impact that would have. I typically contribute on Mondays, but I've heard that in bear markets Fridays may be better, so now I'm doing 1/2 on Mon + 1/2 on Fri.

[Chicken Little]

10 year treasury was as low as 2.25% during GFC

30 year treasury low was 2.69%

I hate to say it but i feel negative rates will be coming.

VJ

That would be a good result for the strategy, correct?

If you had to rank practically poor scenarios, and inflation is ruled out, what would be next?

Something like a steady decline in stocks without much of a move down in treasury yields, over an extended period?

[BayStater]

Maybe this has been posted, but there are a lot of pages and despite the shelter-in-place, I really have better things to do. Is there a way to back test to determine what are the historically best days/months/etc. to contribute & re-balance a portfolio? Would be curious what impact that would have. I typically contribute on Mondays, but I've heard that in bear markets Fridays may be better, so now I'm doing 1/2 on Mon + 1/2 on Fri.

You could solve this as an optimization problem, sure. But in the process, you'd likely just be overfitting the data. I'd assume there's likely not enough evidence to show a significant pattern (the third Tuesday of the second month of each quarter), and any pattern you do see may not continue into the future.

[MotoTrojan]

People say that higher risk means higher reward. That's wrong. Higher risk is a chance of both higher rewards and higher losses.

This is not said often enough.

In any case where one is taking smart risk (no idiosyncratic risk), I would say higher expected reward is fair too.

[MotoTrojan]

Maybe this has been posted, but there are a lot of pages and despite the shelter-in-place, I really have better things to do. Is there a way to back test to determine what are the historically best days/months/etc. to contribute & re-balance a portfolio? Would be curious what impact that would have. I typically contribute on Mondays, but I've heard that in bear markets Fridays may be better, so now I'm doing 1/2 on Mon + 1/2 on Fri.

The only answer with any true fundamental backing is to invest as soon as you possibly can (when the cash clears your account), regardless of day, condition, or which NFL conference won the SuperBowl.

[LittleBitMore]

Maybe this has been posted, but there are a lot of pages and despite the shelter-in-place, I really have better things to do. Is there a way to back test to determine what are the historically best days/months/etc. to contribute & re-balance a portfolio? Would be curious what impact that would have. I typically contribute on Mondays, but I've heard that in bear markets Fridays may be better, so now I'm doing 1/2 on Mon + 1/2 on Fri.

The only answer with any true fundamental backing is to invest as soon as you possibly can (when the cash clears your account), regardless of day, condition, or which NFL conference won the SuperBowl.

What about only investing in the last 2 years of the president's term? I can see a fundamental reason there and the returns are quite stunning

https://gbr.pepperdine.edu/2010/08/pres ... et-cycles/

[LittleBitMore]

10 year treasury was as low as 2.25% during GFC

30 year treasury low was 2.69%

I hate to say it but i feel negative rates will be coming.

VJ

That would be a good result for the strategy, correct?

If you had to rank practically poor scenarios, and inflation is ruled out, what would be next?

Something like a steady decline in stocks without much of a move down in treasury yields, over an extended period?

Conceptually negative rates would be good but the FED has been very clear they don't believe in negative rates (e.g. last speech Powell described us at "the effective lower bound" for rates). So them breaking on that could lead to some panic (i.e. volatility while trending down)

To me the next risk is the relative decline of the US as an economic power -- absent that, US large caps should have strong performance & LTTs will be a safe haven. For example, if we handle COVID19 particularly poorly vs everyone else, we'll spend trillions fighting that rather than reinvesting in growth, and a prolonged recession could set us back.

Alternatively, someone recognizing that low interest rates creates systematic moral hazard (e.g. borrow cheap and buyback stocks until any crisis risks bankruptcy) and taking a stand (i.e. pushing rates back up to 3% in the next 5-7 years). That certainly isn't priced into LTTs and would hurt stocks as well.

[perfectuncertainty]

Hedgie - I substituted QLD for UPRO and the returns are similar but with less risk. Here is the backtest:

QLD + TMF (55/45)

• CAGR = 32.44%
  Max Drawdown = -17.36%
  Sortino Ratio = 3.18
  Best Year = 63.5%
  Worst Year = -5.18%
  Market Correlation = 0.34

UPRO + TMF (55/45)

• CAGR = 32.92%
  Max Drawdown = -19.152%
  Sortino Ratio = 2.78
  Best Year = 73.88%
  Worst Year = -14.16%
  Market Correlation = 0.6

Thoughts?

[MotoTrojan]

Maybe this has been posted, but there are a lot of pages and despite the shelter-in-place, I really have better things to do. Is there a way to back test to determine what are the historically best days/months/etc. to contribute & re-balance a portfolio? Would be curious what impact that would have. I typically contribute on Mondays, but I've heard that in bear markets Fridays may be better, so now I'm doing 1/2 on Mon + 1/2 on Fri.

The only answer with any true fundamental backing is to invest as soon as you possibly can (when the cash clears your account), regardless of day, condition, or which NFL conference won the SuperBowl.

What about only investing in the last 2 years of the president's term? I can see a fundamental reason there and the returns are quite stunning

https://gbr.pepperdine.edu/2010/08/pres ... et-cycles/

I would say this is not a strong fundamental reason for continued outperformance personally. It is not risk-based - While there are some that believe there are long-lasting behavioral reasons for increased returns (take the value premium for instance, although that also has a risk-based explanation), a risk-based explanation is IMHO always much stronger (longer-duration bonds having higher yields for example, or corporate vs. treasury spreads). Why couldn't this be arbitraged away?

Even if you showed that there are less downturns during these periods, the idea of sitting out of the market for 2-6 year cycles is a tough hurdle to get over. Now if your target AA was 25% equity, and instead of holding that perpetually you just held 100% equity in these 1-2 year periods (depending on number of terms, which obviously in some cases requires prescience) to maintain a long-term 25% average, then you'd at-least have the same expected return even without any premium, but you'd be reducing a lot of time-diversification. If you wanted higher than 25% then you could leverage during these periods.

[guyinlaw]

UPRO-TMF (55-45) Performance Mar-April 2020

UPRO-TMF -- down 9.16%
SPY -- down 1.35%

With Long term interest rates at its lowest in the near term, TMF will be not provide risk parity IMO. Fed will be holding it "flat" in the near term.

https://www.portfoliovisualizer.com/bac ... tion4_2=55

[Hydromod]

Maybe this has been posted, but there are a lot of pages and despite the shelter-in-place, I really have better things to do. Is there a way to back test to determine what are the historically best days/months/etc. to contribute & re-balance a portfolio? Would be curious what impact that would have. I typically contribute on Mondays, but I've heard that in bear markets Fridays may be better, so now I'm doing 1/2 on Mon + 1/2 on Fri.

You could solve this as an optimization problem, sure. But in the process, you'd likely just be overfitting the data. I'd assume there's likely not enough evidence to show a significant pattern (the third Tuesday of the second month of each quarter), and any pattern you do see may not continue into the future.

The backtesting I did in this thread suggests that around the beginning/end of months and quarters has historically tended to give better results for rebalancing, if one is using the timing approach instead of the band approach.

To a certain extent my backtesting results were sensitive to rebalance timing relative to a relatively few big market events (e.g., if rebalancing triggered before or after Black Monday). That's just luck of the draw IMO, and I don't see how it's actionable.

I did see some outperformance due to rebalance timing relative to the start/end of the month, maybe because the relative movement of equity and bond prices seems to be a little out of phase during the month. I also saw some performance gain with more frequent rebalancing, such as daily or every two days, but that outperformance died off quickly as the rebalancing period increased.

There also appears to have been a slight reduction in volatility in the middle of the week, at least at the tails, but I don't know what to do with that information. Maybe the really big events were clustered towards the beginning and end of the week...

I didn't specifically look at contributions. Others have studied the effect of dollar cost averaging, and have found that 2/3 of the time lumping it in works out better than spreading it out. This would suggest dumping in contributions at the earliest possible moment would tend to give a better average return, because 2/3 of the contributions will do better. Time in the market versus timing the market and all that...

If I was on a monthly schedule for both contributions and rebalancing, probably I'd pick the beginning/end of the month for both (for superstition's sake if nothing else).

[euphoric85]

UPRO-TMF (55-45) Performance Mar-April 2020

UPRO-TMF -- down 9.16%
SPY -- down 1.35%

With Long term interest rates at its lowest in the near term, TMF will be not provide risk parity IMO. Fed will be holding it "flat" in the near term.

https://www.portfoliovisualizer.com/bac ... tion4_2=55

Making any conclusions based on one downturn/2 months of data seems flawed

[KSActuary]

Not flawed as it accurately reflects what will happen if rates remain flat while the markets remain volatile.