SWR - Trinity Study - Annual average returns - Equity and Bond portions

[Wadiman]

Hi all -

I currently have around 70% equities and 25% bonds/5% cash in my FIRE portfolio which I am now starting to draw down as I FIREd late 2022.

For the bond portion I hold direct corporate (generally investment-grade) bonds and have an overall yield to maturity of circa 6%.  I intend to hold each issue until it matures so pretty confident that I'll get back the principal and all the coupon payments going through to maturity.

This has got me wondering about the assumptions/data used to determine the SWRs in the Trinity Study and whether or not I should take advantage of current high bond yields and lock-in some additional funds and increase the bond allocation to say 35 - 40%.

Does anyone know what the long-term equity and bond returns were that informed the 4% SWR guide?

Thanks!

[MustacheAndaHalf]

Decades ago people used general ideas like "4% SWR rule", but now better tools are available online for free.  Vanguard's nest egg calculator lets you plug in your allocation (70% stock / 25% bond / 5% cash) and run simulations over 30 years, and report what percentage hit $0. 

Vanguard's simulation shows a 4% withdrawal has a success rate of 90%.  Is a 10% chance of hitting $0 acceptable?
What if you live on 3.5% over 30 years - then success is 94% according to this simulation.
https://www.vanguard.com/nesteggcalculator

Note with any simulation, future stock returns may differ from past returns, fooling the simulation.  I would encourage you to use the most recent tools available, instead of relying on "4% SWR".

[Wadiman]

Thanks for that Mustache + 1/2 -

I've run simulations and happy with the results but as far as I undertstand they are based on bond funds and not direct holdings so not sure if comparable. 

Have just googled 30 yr equities performance and can see the return for 1992 - 2021 is 9.9% (total US market that is) so there's still quite a gulf between that and the bond portfolio return so I guess that answers my question.

[MustacheAndaHalf]

If I can pose a question back to you - what term and credit risk are you taking with corporate bonds?

I have accounts at Vanguard, Schwab and IBKR where I could buy corporate bonds, but I have $0 room to buy bonds in my IRAs (which are like Australia's supers).

For the bond portion I hold direct corporate (generally investment-grade) bonds and have an overall yield to maturity of circa 6%.  I intend to hold each issue until it matures so pretty confident that I'll get back the principal and all the coupon payments going through to maturity.

[Wadiman]

The bonds are almost all investment grade (Standard and poors or Moody ratings).  Terms range from 2 - 7 years.  Yield to maturities range from 4.5 to 6.5%

[vand]

You are not really asking the right questions - if it were just a matter of the average return then the whole SWR question would be simple, but the sequencing of those returns is far more important, and as it is nigh impossible to predict what markets will do over the short and even medium term then you are left in the same boat as everyone else without a crystal ball.

[maizefolk]

Does anyone know what the long-term equity and bond returns were that informed the 4% SWR guide?

Equities are typically "US equities" that consist of the historical S&P 500 data going back as far as it can reasonably be traced and then substituting homemade indices of major stocks in the USA prior to that.

Bonds are typically the US federal 10 year bond.

The most common dataset used these days for 4% rule like calculations is the Shiller dataset which goes from 1871-present for both of these datasets.

http://www.econ.yale.edu/~shiller/data.htm

[SeattleCPA]

I think a couple of other points worth noting.

First, the retirement start dates that failed (and which therefore "set" that SWR based on sequence of returns return) are the ones that start in 1966, 1967, 1968 and 1969. (I've just been blogging about this here: Working Longer Avoids Sequence of Returns Risk.)

If you look at those bad early years that set the SWR, it's not just the flat equity markets or the nifty fifty crashing, it's the stagflation of the 1970s. Thus, I don't think going out and getting attractive-looking "high" nominal yields is necessarily comforting at all. A 7% nominal interest rate over the time frame where inflation runs 5% is equivalent to a 2% real interest rate.

The other thing I'd say--and I'm a big fan of bonds don't get me wrong--but choosing bonds means you slightly reduce your downside risk. So with bonds in your portfolio, you maybe make the worst-case scenario a little worse. That sounds great. But the cost is, probably, accepting a range of outcomes that are significantly more modest. You give up all the best case scenarios for example. And you also probably end up with a slightly lower outcome.

All that's a long way of saying that even though I like bonds? (I use David Swensen's asset allocation and so have 30 percent in Treasuries.) I kinda think bonds aren't the right medicine here.

[MustacheAndaHalf]

The bonds are almost all investment grade (Standard and poors or Moody ratings).  Terms range from 2 - 7 years.  Yield to maturities range from 4.5 to 6.5%

I think that is considered short term corporate bonds.

I guess I wasn't clear in mentioning "credit risk", and should have said credit ratings of the bonds.  For example, AAA / AA / A / BBB (all investment grade).  Bonds rated "BBB" have a higher risk of default than "AAA" rated bonds.

When I searched for corporate bond ETFs with only "AAA" / "AA" ratings, I couldn't find any.  In every case, bond ETFs weight heavily towards "A" / "BBB" ratings, almost like they're trying to push the limits of investment grade (until it breaks?).

[MustacheAndaHalf]

(I use David Swensen's asset allocation and so have 30 percent in Treasuries.)

David Swensen ran the Yale Endowment until he passed away less than 2 years ago, so that might be a more up to date measure of his views.  It looks like 15% cash & bonds right now.
https://investments.yale.edu/about-the-yio

[reeshau]

Read it for yourself

Bill Bengen's original study:
http://www.retailinvestor.org/pdf/Bengen1.pdf

Trinity Study:
https://www.aaii.com/files/pdf/6794_retirement-savings-choosing-a-withdrawal-rate-that-is-sustainable.pdf

Trinity Study update, by Wade Pfau:
https://web.archive.org/web/20110708072619/http://wpfau.blogspot.com/2010/10/trinity-study-retirement-withdrawal.html

[SeattleCPA]

(I use David Swensen's asset allocation and so have 30 percent in Treasuries.)

David Swensen ran the Yale Endowment until he passed away less than 2 years ago, so that might be a more up to date measure of his views.  It looks like 15% cash & bonds right now.
https://investments.yale.edu/about-the-yio

So Swensen had two portfolio construction approaches. One used alternative asset classes so only a tiny portion of traditional asset classes like public company stocks and bonds. And that's what Yale would presumably still be using. He talks about how to do this in another great book on investing, Pioneering Portfolio Management. (A book I'm now looking for on my office bookcase but which has mysteriously disappeared.)

And then for individuals, his approach was as described in his classic Unconventional Success, which is basically a 70% equities and 30% bonds passive portfolio (Bonds only being treasuries.)

BTW, the way Yale beat the market massively for decades was it always used illiquid alternative asset class investments.

[MustacheAndaHalf]

(I use David Swensen's asset allocation and so have 30 percent in Treasuries.)

David Swensen ran the Yale Endowment until he passed away less than 2 years ago, so that might be a more up to date measure of his views.  It looks like 15% cash & bonds right now.
https://investments.yale.edu/about-the-yio

So Swensen had two portfolio construction approaches. One used alternative asset classes so only a tiny portion of traditional asset classes like public company stocks and bonds. And that's what Yale would presumably still be using. He talks about how to do this in another great book on investing, Pioneering Portfolio Management. (A book I'm now looking for on my office bookcase but which has mysteriously disappeared.)

I had to donate Swensen's books to a library when I moved abroad, but have read them many years ago.  Robert Schiller (a Nobel Prize winner for Economics) taught a course at Yale, and had David Swensen as a guest speaker:
https://www.youtube.com/watch?v=wRdx7kVNQ_E

Most individual investors will have the most overlap with Yale's Endowment in the foreign equities and cash/bonds categories.  I fortunately meet the SEC definition of "Qualified Client", and have been able to invest in hedge funds through a high NW forum I learned about here.  Swensen would probably warn that retail investors can't expect the returns Yale receives from hiring the best fund managers.  But among retail investors, in my opinion, group evaluation of deals is better than going solo.  At any rate, I invest in Venture Capital (top Yale category) and Private Equity (the non-Leveraged sibling of another top Yale category: Leveraged Buyouts).
https://www.sec.gov/rules/final/2021/ia-5904-fact-sheet.pdf
https://forum.mrmoneymustache.com/post-fire/long-angle-hnw-forum/

[SeattleCPA]

(I use David Swensen's asset allocation and so have 30 percent in Treasuries.)

David Swensen ran the Yale Endowment until he passed away less than 2 years ago, so that might be a more up to date measure of his views.  It looks like 15% cash & bonds right now.
https://investments.yale.edu/about-the-yio

So Swensen had two portfolio construction approaches. One used alternative asset classes so only a tiny portion of traditional asset classes like public company stocks and bonds. And that's what Yale would presumably still be using. He talks about how to do this in another great book on investing, Pioneering Portfolio Management. (A book I'm now looking for on my office bookcase but which has mysteriously disappeared.)

I had to donate Swensen's books to a library when I moved abroad, but have read them many years ago.  Robert Schiller (a Nobel Prize winner for Economics) taught a course at Yale, and had David Swensen as a guest speaker:
https://www.youtube.com/watch?v=wRdx7kVNQ_E

Most individual investors will have the most overlap with Yale's Endowment in the foreign equities and cash/bonds categories.  I fortunately meet the SEC definition of "Qualified Client", and have been able to invest in hedge funds through a high NW forum I learned about here.  Swensen would probably warn that retail investors can't expect the returns Yale receives from hiring the best fund managers.  But among retail investors, in my opinion, group evaluation of deals is better than going solo.  At any rate, I invest in Venture Capital (top Yale category) and Private Equity (the non-Leveraged sibling of another top Yale category: Leveraged Buyouts).
https://www.sec.gov/rules/final/2021/ia-5904-fact-sheet.pdf
https://forum.mrmoneymustache.com/post-fire/long-angle-hnw-forum/

Swensen's book, "Pioneering Portfolio Management," is basically about how endowment funds and maybe the very largest family offices need to operate in order to succeed with alternative asset class investing. Great book. (I really need to find out where my copy went.)

One of this observations, which I discussed here, "Successful Active Investor Tips,", is you need to identify top quartile managers in order to get good results.

Note: We have done a lot of hedge fund investor tax returns over the years. And one thing I'll say is, if someone has investments that kick out 50 or 100 page K-1s, you need to budget for giant invoices from your CPA.

[ChpBstrd]

The simple answer which gets to the OP’s point is that:

1) Stocks are currently priced above their long-term mean (current PE=21.8 vs a mean of 17). https://www.multpl.com/s-p-500-pe-ratio

2) Bond yields, I.e. using the 10y treasury yield as a proxy, are getting better but are still below their long-term mean (current yield=3.97% vs a mean of 4.49%). https://www.multpl.com/10-year-treasury-rate

The half-percent gap between treasury yields and the mean historical treasury yield is smaller than the 1.3% gap between the stock earnings yield (1/PE ratio) vs its mean. So on the basis of that alone, bonds seem like the better deal. The plethora of recession signals and the history of rate hikes leading to recession seal the deal in favor of bonds, IMO.

Plus, if you are in your first 5 years of retirement, you should be doing a bond tent (over-allocating to bonds around the retirement date, and then tapering down each year post-retirement). A 60/40 portfolio should be considered on the aggressive side in your shoes. Instead you’re holding an aggressive portfolio.

The Trinity study was done in the 1990’s using a data set from when stocks were cheaper and bonds generally yielded more. It stands to reason that the odds of portfolio success are lower today. I suggest hiding out in treasuries and waiting for a rebalance opportunity in a future recession.

[MustacheAndaHalf]

Swensen would probably warn that retail investors can't expect the returns Yale receives from hiring the best fund managers.  But among retail investors, in my opinion, group evaluation of deals is better than going solo.

One of this observations, which I discussed here, "Successful Active Investor Tips,", is you need to identify top quartile managers in order to get good results.

Note: We have done a lot of hedge fund investor tax returns over the years. And one thing I'll say is, if someone has investments that kick out 50 or 100 page K-1s, you need to budget for giant invoices from your CPA.

According to a July 2022 paper, 14% of retail hedge funds have significant (positive) alpha, while the worst 6% have significant negative alpha.  Overall, retail hedge funds (individuals and family offices) have similar performance to institutional hedge funds.  So it's possible some of Mr Swensen's lessons carry over.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4183411

I can only speculate how other retail investors invest in small, new hedge funds.  I would guess they invest in ~3 a year out of maybe ~10 funds.  I don't know exact data for Long Angle, but suspect they evaluate an order of magnitude more retail hedge funds by people with relevant industry knowledge.  If I'm correct, this will give me an edge over similar retail investors, and perhaps land me in the top 1/4th Swensen recommends.  I have very preliminary evidence I might have skill picking hedge funds (with very low confidence for lack of data & history).

(I'll create a new thread in "Taxes" to ask about K-1s, thanks)

[SeattleCPA]

The simple answer which gets to the OP’s point is that:

1) Stocks are currently priced above their long-term mean (current PE=21.8 vs a mean of 17). https://www.multpl.com/s-p-500-pe-ratio

2) Bond yields, I.e. using the 10y treasury yield as a proxy, are getting better but are still below their long-term mean (current yield=3.97% vs a mean of 4.49%). https://www.multpl.com/10-year-treasury-rate

The half-percent gap between treasury yields and the mean historical treasury yield is smaller than the 1.3% gap between the stock earnings yield (1/PE ratio) vs its mean. So on the basis of that alone, bonds seem like the better deal. The plethora of recession signals and the history of rate hikes leading to recession seal the deal in favor of bonds, IMO.

Plus, if you are in your first 5 years of retirement, you should be doing a bond tent (over-allocating to bonds around the retirement date, and then tapering down each year post-retirement). A 60/40 portfolio should be considered on the aggressive side in your shoes. Instead you’re holding an aggressive portfolio.

The Trinity study was done in the 1990’s using a data set from when stocks were cheaper and bonds generally yielded more. It stands to reason that the odds of portfolio success are lower today. I suggest hiding out in treasuries and waiting for a rebalance opportunity in a future recession.

I don't think there's a simple answer. Investors' ability to time or beat the market isn't any easier now than it usually is.

What I think is that bonds minimize your downside risk a meaningful amount. But you give up a little bit of return on average and you also give up the big upside.

Tip (not to @ChpBstrd because she or he knows but to others interested in exploring effect of bonds on plans): Use Firecalc to explore different bond allocations: 0 percent, 20 percent, 30 percent etc. And pay attention to the average ending balance, the low balance and the high balance at various bond percentages.

What you'll see probably is there are really good reasons to have bonds in your portfolio. Your bonds probably (?) make the worst-case scenario less "bad." But then you'll also see why people don't like bonds because on average they degrade performance. And because you don't get the big win.

[SeattleCPA]

Swensen would probably warn that retail investors can't expect the returns Yale receives from hiring the best fund managers.  But among retail investors, in my opinion, group evaluation of deals is better than going solo.

One of this observations, which I discussed here, "Successful Active Investor Tips,", is you need to identify top quartile managers in order to get good results.

Note: We have done a lot of hedge fund investor tax returns over the years. And one thing I'll say is, if someone has investments that kick out 50 or 100 page K-1s, you need to budget for giant invoices from your CPA.

According to a July 2022 paper, 14% of retail hedge funds have significant (positive) alpha, while the worst 6% have significant negative alpha.  Overall, retail hedge funds (individuals and family offices) have similar performance to institutional hedge funds.  So it's possible some of Mr Swensen's lessons carry over.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4183411

I can only speculate how other retail investors invest in small, new hedge funds.  I would guess they invest in ~3 a year out of maybe ~10 funds.  I don't know exact data for Long Angle, but suspect they evaluate an order of magnitude more retail hedge funds by people with relevant industry knowledge.  If I'm correct, this will give me an edge over similar retail investors, and perhaps land me in the top 1/4th Swensen recommends.  I have very preliminary evidence I might have skill picking hedge funds (with very low confidence for lack of data & history).

(I'll create a new thread in "Taxes" to ask about K-1s, thanks)

The "Pioneering" book really may be worth reading. Or reading again. What I think, and I have seen people invest in this over years or decades I guess, is small investor don't get into very good funds.

Swensen in "Pioneering" says that the Yale formula won't work for most other endowment funds because the investments offices can't do what Yale does and will never get access to the best fund managers. He doesn't say it explicitly, but you come away understanding that Yale, Harvard, the family offices for the Gates family, the Bezos family, the Musk family etc vacuum up all the best fund managers. Thus, what everybody else (other elite private university endowment funds, mere billionaires and centimillionaires) is left with is the stuff outside that top tier.

[MustacheAndaHalf]

Swensen would probably warn that retail investors can't expect the returns Yale receives from hiring the best fund managers.  But among retail investors, in my opinion, group evaluation of deals is better than going solo.

One of this observations, which I discussed here, "Successful Active Investor Tips,", is you need to identify top quartile managers in order to get good results.

Note: We have done a lot of hedge fund investor tax returns over the years. And one thing I'll say is, if someone has investments that kick out 50 or 100 page K-1s, you need to budget for giant invoices from your CPA.

According to a July 2022 paper, 14% of retail hedge funds have significant (positive) alpha, while the worst 6% have significant negative alpha.  Overall, retail hedge funds (individuals and family offices) have similar performance to institutional hedge funds.  So it's possible some of Mr Swensen's lessons carry over.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4183411

I can only speculate how other retail investors invest in small, new hedge funds.  I would guess they invest in ~3 a year out of maybe ~10 funds.  I don't know exact data for Long Angle, but suspect they evaluate an order of magnitude more retail hedge funds by people with relevant industry knowledge.  If I'm correct, this will give me an edge over similar retail investors, and perhaps land me in the top 1/4th Swensen recommends.  I have very preliminary evidence I might have skill picking hedge funds (with very low confidence for lack of data & history).

(I'll create a new thread in "Taxes" to ask about K-1s, thanks)

The "Pioneering" book really may be worth reading. Or reading again. What I think, and I have seen people invest in this over years or decades I guess, is small investor don't get into very good funds.

Swensen in "Pioneering" says that the Yale formula won't work for most other endowment funds because the investments offices can't do what Yale does and will never get access to the best fund managers. He doesn't say it explicitly, but you come away understanding that Yale, Harvard, the family offices for the Gates family, the Bezos family, the Musk family etc vacuum up all the best fund managers. Thus, what everybody else (other elite private university endowment funds, mere billionaires and centimillionaires) is left with is the stuff outside that top tier.

I have read the book, but you don't seem to be replying to my point.  If Yale takes "all the best fund managers", why do 1/7th of retail hedge funds have significant positive alpha (per paper I cited)?
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4183411

[SeattleCPA]

Not trying to be snarky and hope this doesn't come across that way. But wouldn't some set of funds in any sample show they've generated alpha?

This quote from page 13 of the PDF you link to.

The average retail hedge fund does not produce alpha, but in the cross-section, many retail
funds produce positive alpha.

That doesn't seem very compelling.

Neither to me does this comment from the conclusion:

14.3% of retail funds produce alpha,

Isn't that the same thing as saying nearly 86 percent of retail hedge funds fail to produce alpha?

The sample size seems pretty small too.

To avoid these problems, we start our analysis in March 2012 when we first have at least 100 retail and 100 institutional funds.

Does that mean we're talking about 14 retail funds producing good alpha?

Finally, I'm also a little unsure of how they've accounted for the costs. They say they're considering costs early on. And then say they're looking at net returns and considering AUM fees. But I can also read (well skim) the paper and conclude they're really looking just at alpha?

The other thing I'd be cautious about in terms of costs if I were talking with a client about this: Are all the costs of the hedge fund investment really included? E.g., the hedge fund manager's cost may be included. But is the "wealth advisor" or the CPA now needed to do the annual tax return?

[Must_ache]

You are not really asking the right questions - if it were just a matter of the average return then the whole SWR question would be simple, but the sequencing of those returns is far more important, and as it is nigh impossible to predict what markets will do over the short and even medium term then you are left in the same boat as everyone else without a crystal ball.

When I hear "sequence of returns" I think "randomness" which isn't necessarily what that means, but I don't think that is the worry so much as "how expensive is the market".  ChpBstrd's insightful observation that stocks are currently priced above their long-term mean (current PE=21.8 vs a mean of 17) suggests there could be another 22% downside to get back to historical averages.  If I'm hoping to get 6%/yr on the stock market and every two years I get +15% followed by -2% that variability adds almost no risk to my retirement.  If I think the market is that overvalued, maybe I should be aiming for an a retirement number that absrobs a 22% drop.

I think the bigger worry is assets are still inflated.  Let me cherry pick two 14-yr periods (and I'm excluding dividends)
From March 1995 to March 2009 the S&P went up about 4.0%/yr.
From March 2009 to March 2023 the S&P went up about 11.0%/yr.

So long-term should you expect:
A) 11.0% this is the new normal right?
B) 7.5% this is the 28-yr average.
C) 4.0% this is the actual long-term gain and assets are still massively inflated.

I think the risk is that the market will come down and that the long-term average is a bit lower than what has been seen in the last 10-20 years.  I don't think of it as a sequence of returns but rather lower short-term expectations (e.g. declines).

Personally, after years of 100% equities I have moved to about 15% treasuries and plan to continue ramping that up to about 30% as we hear more about where the interest rate is moving.  If the Fed really does get inflation under control in a couple years back into the 2%-3% range, I don't think at age 51 I'll be disappointed to have a lot of short-term stuff at 5%-6% and a lot of medium-term stuff at 4%+  I just don't want to buy all of it at 5% and then have it peak at 6%+

[MustacheAndaHalf]

Not trying to be snarky and hope this doesn't come across that way. But wouldn't some set of funds in any sample show they've generated alpha?

This quote from page 13 of the PDF you link to.

The average retail hedge fund does not produce alpha, but in the cross-section, many retail funds produce positive alpha.

That doesn't seem very compelling.

Neither to me does this comment from the conclusion:

14.3% of retail funds produce alpha,

Isn't that the same thing as saying nearly 86 percent of retail hedge funds fail to produce alpha?

The sample size seems pretty small too.

To avoid these problems, we start our analysis in March 2012 when we first have at least 100 retail and 100 institutional funds.

Does that mean we're talking about 14 retail funds producing good alpha?

Actually I appreciate the questions.  To roughly summarize your view, any sample will have some funds with alpha.  The average retail hedge fund does not have alpha.  Finally, the bottom 86% of retail hedge funds do not produce alpha (note I'm going to address that last point later, near the bottom of this post).

The first two sentences summarize the paper:

We use a novel fund-level measure to identify 877 retail hedge funds. On average, retail funds do not underperform, either on an absolute basis or relative to institutional funds.

In a way, I'm sidestepping your questions - I am saying they could be true for retail hedge funds, but the same questions are true for institutional hedge funds.  Would you agree with that?

A key point is that I don't need to compete with Yale.  If Yale could buy up the alpha from retail hedge funds, they would.  Yet that alpha persists, meaning Yale has not found a way to buy it up.  The summary says alpha remains, and that retail hedge funds do not underperform institutional hedge funds.  We can speculate on the reasons, but it appears Yale cannot buy up the retail hedge fund alpha.

From this I conclude I am not competing with Yale - I don't need to pick better than Yale.  I need to pick retail hedge funds better than other retail hedge fund investors.


One clarification, "at least 100" is a criteria, where you took it to be exactly 100.  The middle of page 7 clarifies:

After removing backfill data, we have a total of 1,038 funds of which 888 are retail funds. The average retail ratio is 43.8% (median 35.5%)

And then "at least 100" appears on page 8, which is followed a few sentences later with:

Our final monthly dataset contains 2,170 total funds (877 retail funds) with an average retail ratio of 18.9%

Which also matches the summary at the start of the paper: close to 900 retail hedge funds.

Finally, I'm also a little unsure of how they've accounted for the costs. They say they're considering costs early on. And then say they're looking at net returns and considering AUM fees. But I can also read (well skim) the paper and conclude they're really looking just at alpha?

The other thing I'd be cautious about in terms of costs if I were talking with a client about this: Are all the costs of the hedge fund investment really included? E.g., the hedge fund manager's cost may be included. But is the "wealth advisor" or the CPA now needed to do the annual tax return?

I assume this paper looked at a platform - somewhere that invites both investors and hedge funds to meet each other.  I assume fees are disclosed, but industry standard seems to be "2 and 20" (2% AUM fee, 20% of profits).  I have not personally explored the platforms they mention in the paper, so I don't know.

I can invest without a wealth advisor, so I assume their fees are not required.  If CPA fees are hourly or per form, they would impact very small investments far more than larger ones.

I think this also provides some context for the paper saying 14% of retail hedge funds had alpha.  I believe "alpha" can have different meanings, but in this context meant relative to hedge fund factors.  For example, private equity funds have historically beaten the S&P 500.  So you could have the average PE fund not have alpha relative to other PE funds, but still have alpha relative to the public markets.  And then you might use hedge fund costs to decide if that alpha persists after costs.

[ChpBstrd]

Swensen would probably warn that retail investors can't expect the returns Yale receives from hiring the best fund managers.  But among retail investors, in my opinion, group evaluation of deals is better than going solo.

One of this observations, which I discussed here, "Successful Active Investor Tips,", is you need to identify top quartile managers in order to get good results.

Note: We have done a lot of hedge fund investor tax returns over the years. And one thing I'll say is, if someone has investments that kick out 50 or 100 page K-1s, you need to budget for giant invoices from your CPA.

According to a July 2022 paper, 14% of retail hedge funds have significant (positive) alpha, while the worst 6% have significant negative alpha.  Overall, retail hedge funds (individuals and family offices) have similar performance to institutional hedge funds.  So it's possible some of Mr Swensen's lessons carry over.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4183411

I can only speculate how other retail investors invest in small, new hedge funds.  I would guess they invest in ~3 a year out of maybe ~10 funds.  I don't know exact data for Long Angle, but suspect they evaluate an order of magnitude more retail hedge funds by people with relevant industry knowledge.  If I'm correct, this will give me an edge over similar retail investors, and perhaps land me in the top 1/4th Swensen recommends.  I have very preliminary evidence I might have skill picking hedge funds (with very low confidence for lack of data & history).

(I'll create a new thread in "Taxes" to ask about K-1s, thanks)

The "Pioneering" book really may be worth reading. Or reading again. What I think, and I have seen people invest in this over years or decades I guess, is small investor don't get into very good funds.

Swensen in "Pioneering" says that the Yale formula won't work for most other endowment funds because the investments offices can't do what Yale does and will never get access to the best fund managers. He doesn't say it explicitly, but you come away understanding that Yale, Harvard, the family offices for the Gates family, the Bezos family, the Musk family etc vacuum up all the best fund managers. Thus, what everybody else (other elite private university endowment funds, mere billionaires and centimillionaires) is left with is the stuff outside that top tier.

I have read the book, but you don't seem to be replying to my point.  If Yale takes "all the best fund managers", why do 1/7th of retail hedge funds have significant positive alpha (per paper I cited)?
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4183411

I wasn't asked, but 1/7th of hedge funds could have significant positive alpha if they were monkeys typing random ticker symbols. Out of hundreds of funds employing thousands of could-be gurus, it would be a statistical oddity if at least some didn't luck into beating the market, and it would be odd if at least some didn't beat the market multiple years in a row. Flipping a coin and getting heads four times in a row may look like it can't be random, but it's equally probable to the HTHT outcome.

Plus performance doesn't really account for the risk they took. I could buy the market's highest-beta stocks and usually beat the market, but only because the market usually goes up. Similarly, I could buy the S&P500 and then sell call options to beat the market whenever there wasn't a banner year. On the flipside, I could buy the S&P500 plus protective puts and beat the market whenever there was a down year.

Each of these strategies is just redistributing performance. The high-beta strategy takes returns from market-down years and moves them to market-up years. The covered call strategy takes returns from high-return years and moves them to low-return years. The protective put strategy takes returns from market-up years and moves them to market-down years. Given that certain patterns can persist for years at a time, there will inevitably be active investors who beat the market for several years in a row.

[MustacheAndaHalf]

Swensen would probably warn that retail investors can't expect the returns Yale receives from hiring the best fund managers.  But among retail investors, in my opinion, group evaluation of deals is better than going solo.

One of this observations, which I discussed here, "Successful Active Investor Tips,", is you need to identify top quartile managers in order to get good results.

Note: We have done a lot of hedge fund investor tax returns over the years. And one thing I'll say is, if someone has investments that kick out 50 or 100 page K-1s, you need to budget for giant invoices from your CPA.

According to a July 2022 paper, 14% of retail hedge funds have significant (positive) alpha, while the worst 6% have significant negative alpha.  Overall, retail hedge funds (individuals and family offices) have similar performance to institutional hedge funds.  So it's possible some of Mr Swensen's lessons carry over.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4183411

I can only speculate how other retail investors invest in small, new hedge funds.  I would guess they invest in ~3 a year out of maybe ~10 funds.  I don't know exact data for Long Angle, but suspect they evaluate an order of magnitude more retail hedge funds by people with relevant industry knowledge.  If I'm correct, this will give me an edge over similar retail investors, and perhaps land me in the top 1/4th Swensen recommends.  I have very preliminary evidence I might have skill picking hedge funds (with very low confidence for lack of data & history).

(I'll create a new thread in "Taxes" to ask about K-1s, thanks)

The "Pioneering" book really may be worth reading. Or reading again. What I think, and I have seen people invest in this over years or decades I guess, is small investor don't get into very good funds.

Swensen in "Pioneering" says that the Yale formula won't work for most other endowment funds because the investments offices can't do what Yale does and will never get access to the best fund managers. He doesn't say it explicitly, but you come away understanding that Yale, Harvard, the family offices for the Gates family, the Bezos family, the Musk family etc vacuum up all the best fund managers. Thus, what everybody else (other elite private university endowment funds, mere billionaires and centimillionaires) is left with is the stuff outside that top tier.

I have read the book, but you don't seem to be replying to my point.  If Yale takes "all the best fund managers", why do 1/7th of retail hedge funds have significant positive alpha (per paper I cited)?
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4183411

I wasn't asked, but 1/7th of hedge funds could have significant positive alpha if they were monkeys typing random ticker symbols. Out of hundreds of funds employing thousands of could-be gurus, it would be a statistical oddity if at least some didn't luck into beating the market, and it would be odd if at least some didn't beat the market multiple years in a row. Flipping a coin and getting heads four times in a row may look like it can't be random, but it's equally probable to the HTHT outcome.

Plus performance doesn't really account for the risk they took. I could buy the market's highest-beta stocks and usually beat the market, but only because the market usually goes up. Similarly, I could buy the S&P500 and then sell call options to beat the market whenever there wasn't a banner year. On the flipside, I could buy the S&P500 plus protective puts and beat the market whenever there was a down year.

Each of these strategies is just redistributing performance. The high-beta strategy takes returns from market-down years and moves them to market-up years. The covered call strategy takes returns from high-return years and moves them to low-return years. The protective put strategy takes returns from market-up years and moves them to market-down years. Given that certain patterns can persist for years at a time, there will inevitably be active investors who beat the market for several years in a row.

I believe alpha of hedge funds is different from alpha relative to the S&P 500.  For example, private equity (PE) funds have historically beaten the S&P 500 going back decades.  Yet within PE funds, some have alpha relative to other hedge funds.  Confusing the two can lead you to believe only some PE funds beat the S&P 500, when the average PE fund actually does.

It would be rare for "hedge funds" to buy "ticker symbols" at all.

You bring up "redistributing performance" which I believe is not relevant to hedge funds.  Venture capitalists fund companies outside of public markets, long before those companies file an IPO.

[Must_ache]

As long as I'm talking about 14-yr time periods, I think market values still have to adjust from departing this reality.  Interest rates were too low for way too long.

[SeattleCPA]

I believe alpha of hedge funds is different from alpha relative to the S&P 500.  For example, private equity (PE) funds have historically beaten the S&P 500 going back decades.  Yet within PE funds, some have alpha relative to other hedge funds.  Confusing the two can lead you to believe only some PE funds beat the S&P 500, when the average PE fund actually does.

So I started to type a response to above comment, which is itself a response to @ChpBstrd 's comment about alpha. And now I'm not sure what the authors mean or what readers understand.

I was going to say that alpha shows whether the fund manager is doing "something" that lets them beat their benchmark. Something more than just investing in riskier assets. That's what I thought. But maybe the term needs to defined a bit more?

FYI, I Googled beta, skimmed through a bunch of layperson definitions from the usual sources, and find myself now aware that my MBA class memory (er, which is decades old) is not among the popular definitions.

@MustacheAndaHalf sorry if I didn't read the paper closely enough and so misunderstood the sample size. I will say that one issue I inadvertently ended up leaving out of my posted response was how short the time period looked. Six, seven years? That seems short. BTW agree with you about the costs.

[SeattleCPA]

When I hear "sequence of returns" I think "randomness" which isn't necessarily what that means, but I don't think that is the worry so much as "how expensive is the market".

For what it's worth, I think the market being expensive is something different than variability.

So for example, maybe historically the arithmetic mean of stocks was 10% and the standard deviation was 15%.

But maybe now the expected arithmetic mean of stocks looks to be 7% but the standard deviation still looks to be 15%.

You and I can (and should IMHO) adjust downward the SWR we use given the lower expected returns (10% vs 7%). But if the standard deviation is still high (still 15%), we've got the same randomness. And the chance of a bad patch of returns as we start retirement is still ever present.

BTW, probably, embedded in my thinking is notion that the market isn't really "high". I.e., I agree PE ratios are above the historical average. But I think it's possible they'll just continue to be high. This is same thing as saying that returns going forward are just going to be lower. And that interest rates have been trending lower as the world has accumulated more capital and its financial system gotten better about allocating that capital.

[FIPurpose]

As long as I'm talking about 14-yr time periods, I think market values still have to adjust from departing this reality.  Interest rates were too low for way too long.

Says who? The general trajectory of the last 40 years has been dropping interest rates overall with seasonal humps. If we're just in a hump, then the fed basically dropping to 0 at the next recession still seems somewhat guaranteed in the next 5 years which is why you don't see long-term bond rates going up.

If 2-3 years from now the economy is still chugging along and the fed is still saying no plans to drop rates, maybe those 10-20 year bonds will start to go up. But everyone is expecting the drop, this year or next.

[MustacheAndaHalf]

@MustacheAndaHalf sorry if I didn't read the paper closely enough and so misunderstood the sample size. I will say that one issue I inadvertently ended up leaving out of my posted response was how short the time period looked. Six, seven years? That seems short. BTW agree with you about the costs.

No worries - the more I read the paper, the less I understand it.  I suspect all of us (me included) previously misread "alpha" to be either public markets (CAPM) or hedge fund alpha (FH8).  Apparently not:

Across each model specification (CAPM, Carhart, FH8), both institutional and retail funds consistently produce insignificant alpha. This suggests that retail funds do not underperform. [PAGE 11]

When the paper says "14.3% of retail hedge funds produce positive and significant alpha", they are talking about a "cross-section".  I don't understand it, so I think this paper is mostly lost on me.

The average retail hedge fund does not produce alpha, but in the cross-section, many retail funds produce positive alpha. [PAGE 12]

With the caveat Seattle CPA mentioned about 6-7 years of data, the following quote also matches longer series data I've seen.  I would add another flaw in the study: retail hedge fund investors are locked in for years (sometimes 10 years), so the idea of moving money between hedge funds doesn't make sense to me.  But if you could, that would beat retail mutual funds.

Furthermore, column 5 provides evidence that, under the equal- and market shareweighting schemes, retail hedge funds outperform retail mutual funds by between 2.4% and 3.6% annually. [PAGE 11]

[Mr Mark]

My 2 cents worth. It's a fool's errand to try to divine a forward looking truly "Safe" WR from the entrails of the past.

Any attempt to forecast the future based on the past with certainty is flawed by definition. However, we know from looking at the past:
- markets have always recovered if (nominal) GDP continues to grow.
- index funds - if low cost - have outperformed managed funds and market timing approaches.
- Equities have provided long term net growth. High quality bonds have provided short-medium term 'ballast' (to quote JL Collins).

Historically, a 3% withdrawal rate would last very much forever. Your beneficiaries will be grateful. Anyone waiting for that 3% is almost certainly unnecessarily delaying FIRE.
OTOH, counting on an 8% withdrawal rate probably won't last forever, and might run out quite quickly.

In between is what will probably happen. But what we will ALL do IRL once FIRE'd is "dynamic withdrawal" and if the market has a multiyear bear market and there's a big recession, we will engage plans B,C,D for cutting expenses and adding cash generation as required.

So, I would caution anyone who thinks "my current expenses are ~30k a year, so with a nest egg of 750K in X/Y% at a 4% SWR I can count on inflation adjusted income sufficient for my needs for-eva! no matter what!"

I'm much more comfortable with leaping to FIRE if there's currently a margin on income required from the stash [3.5%?] and there are many self evident ways to both (a)reduce expenses [ie base expenses are not bare bones but include lots of short term non-essential discretionary spending] and (b)boost income [good skills, ability to do short term work, unaccounted inheritances, SS, can move, rent a spare room, some proven and cash flowing passive income stuff, etc].

I think the flexibility and a sort of 'jujitsu' FIRE concept is much more useful, once your expenses are in the 3.5 - 6% of liquid stash range. Especially if you're young, the goal of keeping learning, keeping doing stuff you like - some of which may be able to make you money - is a key insurance policy.

Just remember, no one knows what a real SWR is going forward.

[SeattleCPA]

- Equities have provided long term net growth. High quality bonds have provided short-medium term 'ballast' (to quote JL Collins).

I don't think the above is quite right. Equities provide the best (only?) opportunity for growth (compound returns?). But for individual investors, equities don't always provide the best result. This is one of the fundamental errors that I believe exists in the thinking of JL Collins.

Further, calling bonds "ballast" is too vague IMHO. What does that even mean? Far better I think to understand that the math and history show you dial down your downside risk by adding bonds to your portfolio. But at a cost. You'll almost surely end up with less money. And you give up the big upside.

Tip for anyone that cares: You can use something like FireCalc or cFireSim to play with the bond allocation.

Quoting from Firecalc, their standard inputs with a 25% allocation to bonds produces this result:

The lowest and highest portfolio balance at the end of your retirement was $-300,739 to $4,259,606, with an average at the end of $1,419,766.

And now another quote from FireCalc... their standard inputs with zero allocation to bonds produces this result:

The lowest and highest portfolio balance at the end of your retirement was $-698,263 to $6,381,973, with an average at the end of $2,072,093.

[joemandadman189]

PTF