David Swensen Porfolio

[401k Man]

Hello everyone,

Many years ago I set up a portfolio in alignment with advice offered by David Swensen (RIP) and haven't tinked w/it beyond a quarterly (maybe an occasional twice a year) rebalance. Note: One exception where I split a recommendation into two funds w/the same purpose.)

Over the years this particular account never showed a long-term growth percentage like the other accounts, and I didn't think much of it because it was David Swensen. Last night I decide to dig into each account to discover that the Swensen portfolio remains the lowest performer at at. 5.6% ~ and I have this account set up maybe 15 years ago...and it is about 50% of retirement. Saddened and need to course correct. 5.6% on 10 year view, where other accounts are north of 9.6%.

Yale's Money Guru Shares Wisdom with Masses - NPR
https://www.npr.org/2006/10/05/6203264/ ... ith-masses

Curious if anyone else set up a similar plan? If so, how has it been working for you? Hoping I'm missing something obvious in viewing or my understanding (where the real numbers are stronger than what Amazon Vanguard displays as 5.6%)

[iceport]

Curious if anyone else set up a similar plan? If so, how has it been working for you? Hoping I'm missing something obvious in viewing or my understanding (where the real numbers are stronger than what Amazon displays as 5.6%)

I set up a portfolio that was not inspired by Swensen, but ended up being very similar, though with less TIPS, less REITs, and other less significant differences.

Have you made sure that you are comparing the performance of the Swensen portfolio to the other(s) over exactly the same time period? Also, I'm not familiar with obtaining investing data from Amazon. How are you obtaining the 10 year performance data?

Oh — and one obvious source of error would be if you've been contributing to or withdrawing from the portfolio *at all* during the 10 year period. The timing of any account-level activity would certainly affect the performance. Even the timing of any re-balancing could affect the results.

Finally, to be of better help, we'll need to know your target asset allocation — preferably with the specific funds you are using and the specific percentages of the portfolio they represent — so that the 10 year (time-weighted) performance can be approximated independently.

[401k Man]

Amazon = Vanguard (multi-tasking/corrected)

Portfolio was created all at the same time (say all in the same day.) Rebalanced once a quarter (once in a while twice a year.) Only modification was to break one or two of the categories into two funds instead of one, same category where the category total would equal the Swensen model. Still unclear why the Vanguard dashboard and fund overview wouldn't (isn't) showing the performance of that particular account but is all the others...and
------------------------------------------ Asset Class ----------Category -----------Target
VAIPX.-----Inflation Protected Securities-------- Inflation Indexed ------Inflation-Bond-------15%
VEMAX-----Emerging Markets-----------------Intern/Global---------For.Emerging---------05%
VGSLX-----Real Estate Index------------------Sector----------------Real Estate-----------20%
VIMAX-----Mid-Cap Blend--------------------Domestic-------------Blend - Mid----------08%
VSIAX-----Small-Cap Value-------------------Domestic-------------Value - Small---------08%
VTIAX-----Total International Stock ------------Intern/Global---------Blend - For. Lgv-------03%
VTSAX-----Total Stock Market-----------------Domestic-------------Blend - Large---------15%
VUSTX.-----Long-Term Treasury---------------Long-term Bond-------Long Govt------------15%
VWICX-----International Core Stock-------------Intern/Global---------Blend - For. Lg--------12%

[Nahtanoj]

One of Swensen's guidelines from his book for individual investors ("Unconventional Success") is that no asset class should exceed 30% of the total portfolio. So, for example, his sample portfolio limits US stocks to 30% of the total - very similar to what you have.

Over the last 15 years, US stocks have outperformed virtually everything else, and lots of portfolios have more than 30% in US stocks. If you are comparing yourself to any of those portfolios, it's not surprising that they have outperformed a more diversified portfolio. (Didn't someone say that diversification works, "even when you don't want it to" - meaning, even when it produces a lower return than a less diversified portfolio?)

Also, within US stocks, it looks like you overweighted US small and value stocks, which have underperformed the broad US stock market over much of the last 15 years. I do the same thing, but Swensen doesn't. He just sticks to the total US stock market - market cap weighted, and not tilted to small or value -- and the tilt to small and value would have slightly reduced returns over the last 15 years.

I have been generally using a Swensen-style portfolio structure myself for the last 15 years or so. The six asset classes he recommends for long-term portfolios make sense to me, as do his other principles, including among others:

* maintaining an equity orientation;
* rebalancing regularly;
* having some assets that may help overcome inflation (such as TIPS and real estate);
* using nominal Treasuries to provide some protection against deflation, financial panics, recessions and depressions;
* limiting foreign currency exposure to a maximum of 25 to 30% of a portfolio; and
* limiting the size of any one asset class to not more than 30%.

Since these principles seem pretty persuasive to me, I haven't spent a lot of time benchmarking my portfolio to other portfolio structures, although I am well aware that other portfolios may have performed better over the last 15 years. (For example, a 60/40 portfolio of US large cap stocks and US bonds.)

[telemark]

I've been running a slightly simplified version of that in my Roth IRA, currently at Fidelity using

15% FIPDX (intermediate TIPS)
15% FNBGX (long Treasuries)
30% FSKAX (total stock market)
20% FSRNX (REIT)
20% FTIHX (total international)

I don't break out developed and emerging markets, just use total international for both. Before retiring I ran the same allocation at Vanguard using their funds, but when I started doing Roth conversions it was more convenient to keep everything at
Fidelity.

Using Fidelity's calculations, I see a five-year return of 6.35%, a three-year return of 1.73%, and a one-year return of 10.69%. Breaking it down by year shows

2019 22.41%
2020 -11.84%
2021 20.52%
2022 -20.02%
2023 14.41%
2024 8.35%

Just looking at that I'd guess that the interest rate increases took a toll in 2022, and the pandemic hurt everything in 2020. The volatility is higher than I had expected, but I run a more conservative allocation in my traditional IRA.

A few comments on your allocation:

Using VEMAX and VTIAX means you're including emerging markets twice. And VWICX has an expense ratio of 0.48% -- that's way more than I would pay for a fund, especially at Vanguard. Similarly, breaking out small and mid-cap and then using total stock market means including the small and mid-cap funds twice. Since both of those have lagged behind large cap that probably hasn't helped your returns.

As to whether you should change, it depends on what you are trying to accomplish.

[breakfastinbed]

1) A backtest over the past 25 years suggests this portfolio outgrew a 60/40, but not by a lot. It did grow less than a 70/30. If you add the corresponding international component, however, (50/20/30), Swensen's portfolio did outperform that.

2) This is a common criticism of risk parity style portfolios - they underperformed the total market, or they had low growth - but you're looking at a portfolio that's supposed to be primarily handling decumulation, not accumulation. You shouldn't be looking for this portfolio to grow a ton; it's like asking the towing capacity of a Corvette. How about throwing a 4% annual drawdown, inflation-adjusted, at it?



3) FYI, if you backtest your portfolio (which is limited because you use a fund, VWICX, that is only 5 years old) against the recommended Swensen portfolio, you underperformed it by 1.23% annually (18%), mostly due to your replacement of half of the total market allocation with VIMAX + VSIAX, which together underperformed VTSAX by 4.23% annually (28%). Using an older international index to replace VWICX improves your 15-year performance to only 0.4% less annualized.

4) The other issue (which is more with the base portfolio/recent history) is that 15% each in TIPS and treasuries hurt your return, as both of these had a negative real return over the past 15 years.


Overall, over the past 25 years, Swensen's portfolio outperformed a similar three-fund portfolio of 50% total US, 20% total international, and 30% total bond by 0.48% real annually, with better risk-adjusted returns as well. If you want a higher return going forward, I'd recommend considering adjustments to this portfolio, which was published, I believe, 20 years ago.

[iceport]

Over the years this particular account never showed a long-term growth percentage like the other accounts, and I didn't think much of it because it was David Swensen. Last night I decide to dig into each account to discover that the Swensen portfolio remains the lowest performer at at. 5.6% ~ and I have this account set up maybe 15 years ago...and it is about 50% of retirement. Saddened and need to course correct. 5.6% on 10 year view, where other accounts are north of 9.6%.

Thanks for your portfolio details. I ran them (mostly) through Portfolio Visualizer and came up with a 10 year annualized return of about 5.7%, so it seems to confirm the Vanguard number. To get around the short history of VWICX I just substituted it with VTMGX. It's an imperfect substitution, but I thought it was better to get at least a 10-year return.

Portfolio Visualizer had a stock analysis for the Swensen portfolio, and the 10-year return for that was 6.1%, but they also used long term treasuries. Substituting intermediate term treasuries — which I do believe is more consistent with Swensen's intentions — produced a 10-year return of 6.5%. So that choice does make a difference. EDITED TO ADD: Note that I just realized I used the "David Swensen Yale Endowment" portfolio as a base instead of the "David Swensen Lazy Portfolio," which both show up on the home page of Portfolio Visualizer. The biggest difference is the LT vs. IT treasuries.

Probable reasons for the discrepancy with your portfolio include these differences: 1) your use of long term treasuries underperformed Swensen's intermediate term treasury average; 2) your small cap value tilt underperformed the total US stock market; and 3) VWICX *over*performed Swensen's developed markets allocation.

I also compared these two versions of the Swensen portfolio to a 70/30 Three Fund Portfolio, with 60/40 US/ex-UX, and the 10 year return was 7.2%. This beat the Swensen portfolios, but not by nearly as much as your other allocations, which you say exceeded 9.6%.

The Swensen portfolio probably underperformed due primarily to the heavy REIT allocation, which significantly underperformed US equities, as did ex-US equities.

All of these backtest results are for the 10 year period for calendar years 2015 through 2024.

Possible reasons for your other "portfolios" outperforming include the following: 1) any account-level contributions or withdrawals in the other accounts, particularly on a sustained, ongoing basis; 2) different time periods measured; 3) more US, less international; 4) larger equity allocation; etc.

In other words, you need to be sure you're comparing apples to apples. For example, if you left the Swensen account alone while you were contributing to the others, you'd essentially be comparing the time-weighted performance of the Swensen portfolio against the dollar-weighted performance of the other accounts. Not a fair comparison. You'd want to run your other allocations through something like portfolio visualizer to get the time-weighted returns.

You can see the full backtesting results HERE. You can also edit the portfolios to do your own comparisons.






Now, as for your feeling the need to correct course, I really can't help you there. (Just take a look at my signature lines!) I'm generally loath to make major changes, and would anticipate some reversion to the mean with just about any reasonable asset allocation. Just make sure your comparisons are valid before you do anything drastic.

EDIT #2: Having said that, I think the most obvious change to make — and one I wouldn't hesitate to make, really — is to ditch the long term treasuries for an intermediate term bond fund. The LT treasuries are more volatile and you really haven't been rewarded for the risk.

[iceport]

2) This is a common criticism of risk parity style portfolios - they underperformed the total market, or they had low growth - but you're looking at a portfolio that's supposed to be primarily handling decumulation, not accumulation. You shouldn't be looking for this portfolio to grow a ton; it's like asking the towing capacity of a Corvette.

I don't know where you got this, about being for decumulation. I thought Swensen intended the portfolio for long term growth in the accumulaton phase, with the 70/30 allocation being his idea of a good balance between growth and preservation for retirement portfolios.

He described other specific and more conservative allocations for funds needed in shorter time frames.

[jdibber]

2) This is a common criticism of risk parity style portfolios - they underperformed the total market, or they had low growth - but you're looking at a portfolio that's supposed to be primarily handling decumulation, not accumulation. You shouldn't be looking for this portfolio to grow a ton; it's like asking the towing capacity of a Corvette.

I don't know where you got this, about being for decumulation. I thought Swensen intended the portfolio for long term growth in the accumulaton phase, with the 70/30 allocation being his idea of a good balance between growth and preservation for retirement portfolios.

He described other specific and more conservative allocations for funds needed in shorter time frames.

Swenson is a top 6 (of 20 common portfolios) in total return since 1970, beating out "Three-Fund" portfolio. go figure
https://portfoliocharts.com/charts/portfolio-matrix/

[401k Man]

All - I am blown away by the responses, and the time and detail it took for those responses. Thank you so much!

Few background notes that I recognize would have been good to include in the original post:
• The fund/s discussed are all part of a Roth IRA
• While possibly a pipe dream, I have been looking for the portfolio to apx. double every 10 years. (meaning ideally AVR north of 9%)
• Timeframe to keep portfolios on the more aggressive side would be about another 10 years, then start shifting to more conservative options.

Swensen portfolio was set up all at one shot (mentioned above) and early on, longest one I have running. I do not recall when I split one or two of the allocations in half nor the reason for the split; speculating that it was either because a similar fund in the same category/class (checked the box) was doing better so split into two, or (more likely) I have been also using Vanguard's portfolio review where it would highlight when a portfolio is getting too high/low in a particular category. That said, I have been staying the course (yes, I do question) and I think any adjustments to map better to Vanguard's recommendations were done through other accounts.

Comparing apples-to-apples. Is the intention to compare apples-to-apples, yes but not a carbon copy understanding variances are expected for areas such as timing. So ideally the comparison would be my version of Swensen to Swensen's version of Swenson. Now, there is a question if Swenen's portfolio would be considered conservative (a point I wouldn't have considered back when I decided to use the model,) resulting in my own expectations (better or worse, and changing.) Comparing this portfolio to other portfolios in the Roth IRA are mainly for a reality check...decided to take a deeper look and was surprised by how it wasn't performing as well over these past exceptional years (yes, in comparison to my other portfolios and market in general.) Maybe part of the concern is that this portfolio represents apx. 40% weight against the other portfolios, so it seems as if it should have done better. Then again, w/countless rabbit holes and considerations, if the world markets take drastic changes this Swensen portfolio starts looking good again. I recognize my own expectations are part of the doubt. To me, it seems returns should have been higher.

I will likely make adjustments (goal, to more closely match Swensen Yale) based on everyone's input.
Can someone further elaboration on the overweight given to us small and value funds (likely unintentional shift)?
Interested in further understanding swapping long term treasuries for an intermediate term bond fund (and how each represent the same category)

Also good point on draw down and such later - I could see this potfolio lasting through retirement when discussing spend down.

[jdibber]

OP....nice little commentary from optimize portfolio. not sure if you've seen it. they do a nice job on that site

https://www.optimizedportfolio.com/davi ... portfolio/

[iceport]

Interested in further understanding swapping long term treasuries for an intermediate term bond fund (and how each represent the same category)

This is the easy one. Swensen recommended treasuries with a range of durations. The article you linked in the OP included three Vanguard funds for short, intermediate, and long term treasury bonds. The intent was to produce a bond portfolio with an overall intermediate term average. In a subsequent article (NPR has featured David Swensen many times over the years), Swensen's treasury bond recommendation is condensed to a single fund, Vanguard Intermediate-Term Treasury Fund Investor Shares (VFITX).

3 Investment Gurus Share Their Model Portfolios

If you're going to use a single treasury bond fund, it should be an intermediate term treasury bond fund.

Here are 6 Vanguard treasury bond funds, short, intermediate, and long term in both active and index fund form, compared to the total bond market index fund. These are all bond funds, so it's all relative, but the only two funds that went on a (relatively) wild ride and ended lower were the two long term treasury bond funds.



If it were me, even if I were trying to duplicate Swensen's model portfolio, I'd be absolutely fine holding any one of the intermediate term bond funds in this chart, including VBTLX. If you want to hold treasuries, I'd recommend just picking VSIGX, Vanguard Intermediate-Term Treasury Index Fund Admiral Shares, and call it a day. The active version is almost as cheap and provides the same exposure.

Also good point on draw down and such later - I could see this potfolio lasting through retirement when discussing spend down.

Again, the Swensen model portfolio was not intended (solely) as a decumulation portfolio.

Adjust Your Portfolio as You Near Retirement: As you get older and closer to retirement, it's obviously important to have enough money in less risky, more predictable investments.

But rather than changing all the numbers on your "asset allocation" pie chart depending on your age, Swensen prefers to think about this question in a way that's easier to grasp. He says as people age, they can keep their investment portfolio set up the way it always has been. But they should start to move money out of it, across all investment categories proportionally, and transfer the money into an account that's invested in money-market funds or short-term, inflation-indexed bonds.

David Swensen's Advice for the Individual Investor

Feel free to disagree with Swensen, but this was his generic advice.

[anoop]

This looks like a 70/30 portfolio. When I see something like this, I think I would do fine with 70% SPY and 30% t-bill without worrying about trying to break things down into international, real estate, etc.

[iceport]

Can someone further elaboration on the overweight given to us small and value funds (likely unintentional shift)?

This is pretty simple, too, qualitatively, without trying to quantify the effects.

The Swensen recommended a broad based "total market" index fund for exposure to US equities. You use one for half of the allocation, but also use separate mid-cap and small-cap value sector funds for the other half. Over the past 10 years, the two sector funds have underperformed the total cap-weighted market. Thus, you have underperformed the model portfolio.



I'm not making a recommendation — as I noted before, I hate making portfolio changes — but there's really not much of a mystery here. If you want to better track the performance of the model portfolio, go back to the model, ditch the two sector funds and put the entire allocation into VTSAX.

But if you do make a change, just realize that your failed attempt* to "improve upon" the model might actually start working in the future!



* This is, admittedly, a presumption on my part. Tilting towards smaller and value companies has often been expected to outperform the total market. Often, maybe even usually, it does. But not always, as these last 10 years have shown clearly.

[jdibber]

minor side note, but REITs are pretty much mid cap blends as well, so there was 28% devoted to midcap blend, along with the overlap found in total US and International funds.

[DesertGator]

One thing I like to suggest people consider is that they understand the objectives of the expert or role model that are "following" for their own retirement investing choices. Swensen was not investing the Yale endowment for retirement - it is a permanent portfolio that could model annual inflows and expect outflows forever. It differs from the objectives of a typical long term retirement investors who thru much of their life is in an accumulation phase, derisking as retirement approaches and through elderhood, and decumulation, perhaps planning explicitly to leave, or not leave an amount for inheritance.

I think Warren Buffet is great, but his choices are not necessarily relevant to most long term retirement investors either, except for the idea that he tends to buy large healthy companies, and these are not typically terrible investments.

One could certainly do worse than following these kinds of folks....

[telemark]

On the duration question, I should add that the long treasuries are only in my Roth. In my traditional IRA, I use a mixture of intermediate treasuries and money market (with MM substituting for short term treasuries), thus diversifying across durations in the overall portfolio. I would not want to use long term for all of my bond allocation -- too volatile.

P.S. It's worth pointing out that in Swensen's book he lays out a series of principles for portfolio construction, and the reasoning behind them. The reference portfolio is one example of how to put those principles into practice, but any number of others can accurately be described as Swensen portfolios.

[iceport]

P.S. It's worth pointing out that in Swensen's book he lays out a series of principles for portfolio construction, and the reasoning behind them. The reference portfolio is one example of how to put those principles into practice, but any number of others can accurately be described as Swensen portfolios.

That may be true, but Swensen joined virtually every single other authoritative advocate of passive investing (for individuals, not endowments) in urging investors to stick with whatever portfolio they end up choosing "through thick and through thin." It's self defeating to continue to tweak an asset allocation in response to recent market performance or Vanguard's blind prodding to diversify.

Basically, his advice for individuals was the polar opposite of the hyperactive trading he used for the Yale endowment.

[telemark]

P.S. It's worth pointing out that in Swensen's book he lays out a series of principles for portfolio construction, and the reasoning behind them. The reference portfolio is one example of how to put those principles into practice, but any number of others can accurately be described as Swensen portfolios.

That may be true, but Swensen joined virtually every single other authoritative advocate of passive investing (for individuals, not endowments) in urging investors to stick with whatever portfolio they end up choosing "through thick and through thin." It's self defeating to continue to tweak an asset allocation in response to recent market performance or Vanguard's blind prodding to diversify.

Basically, his advice for individuals was the polar opposite of the hyperactive trading he used for the Yale endowment.

Oh yes, absolutely. I think you may have mistaken my meaning. And if you wish to verify its truth, I refer you to Chapter 3 of Unconventional Success, particularly the section subtitled The Art Of Personalization.

[iceport]

That may be true, but Swensen joined virtually every single other authoritative advocate of passive investing (for individuals, not endowments) in urging investors to stick with whatever portfolio they end up choosing "through thick and through thin." It's self defeating to continue to tweak an asset allocation in response to recent market performance or Vanguard's blind prodding to diversify.

Basically, his advice for individuals was the polar opposite of the hyperactive trading he used for the Yale endowment.

Oh yes, absolutely. I think you may have mistaken my meaning. And if you wish to verify its truth, I refer you to Chapter 3 of Unconventional Success, particularly the section subtitled The Art Of Personalization.

No, I got you, and I agree with you. My comment was directed to the OP, who seems to be doing a very good job staying the course overall, but might be tempted to tweak around the edges. Every change made, though it might seem like a good idea at the time, is just about as likely to be a drag on future returns as it is to help. Any reasonable asset allocation an investor can stick with over the long term is bound to outperform the elusive "optimum" asset allocation that keeps changing.