Should I Switch to a Three Fund Portfolio

Switch to a Three Fund Portfolio

When I recommended an investment plan for Stacey, I suggested she setup a three fund portfolio.  It’s a simple, straightforward target that will accomplish the investment goals of a vast majority of people.  In fact, it’s my default recommendation to anyone who asks for allocation advice.  And this question has been nagging at me ever since:  Should I switch to a three fund portfolio myself?

My Current Allocation


A few years ago, my wife and I agreed to an Investment Plan.  Part of that plan includes a target for our retirement portfolio investments:

65% Stock  |  30% Bond 5% REIT


For us, REIT simply means a 5% stake in Vanguard Real Estate Index Fund Admiral Shares (VGSLX).  Simply done.

30% Bond

Your bond allocation should make you uncomfortable, and I sure don’t like mine.  I wake up thinking it should be smaller and I go to sleep thinking it should be bigger — so it’s just about right for us.  Originally, we had this split up between domestic and international bond funds.  But I couldn’t justify the international bond fees so this category became simple, too:  30% stake in Vanguard Total Bond Market Index Fund Admiral Shares (VBTLX).

Incidentally, If you have a few different retirement accounts between IRAs, 401(k)s, 403(b)s, 457(b)s, and regular taxable accounts; the ubiquity of VBTLX can really help you when it’s time to rebalance.

65% Stock

Our IPS calls for the stock stock portion of our portfolio to be split 60% Domestic and 40% International.  At the time we created this plan, I had read a lot about how slicing and dicing within asset classes wasn’t that important, but might bring modest returns over an undifferentiated allocation.  I love tracking, and I don’t mind a little bit of complication, so we settled on this target:


It’s really fun to set up a portfolio like this.  I like optimizing holdings between different accounts, and I think there truly is some value (maybe around 0.5% per year) to sticking diligently to an asset allocation like this one.  Principally, this is because I’m forced to “buy low” when a certain asset class (like international stocks in 2018) falls out of favor and “sell high” when an asset class performs well.  Note, I’m not actively buying and selling, just keeping to my target allocation by adjusting contributions or rebalancing not more than once per year!  Inevitably, the high flyers come back to earth and the stragglers catch up.  Sticking to an allocation is a contrarian game that I think produces results.

Did I mention, I really enjoy tracking the components?  I do.  See more about that below.


Oh, are you trying to have fun with your portfolio?  Be careful, there’s a lot of “fun” ways to lose your shirt.  As a rule, good investing should be simple and boring.

Keeping it simple also means you can spend more time doing other stuff; you have fewer things to keep track of and fewer moving parts to distract you.  I’m a bit jealous when I see someone who is just 100% VTSAX.  They don’t spend any time hemming or hawing about where to put their money: it’s fire and forget.

The last downside, which I’ve been thinking about lately, is for my wife and kids.  The reason to buy life insurance is to make sure that my unexpected death would not imperil the livelihood of those who depend on me.  My wife is hard working and smart, but I don’t think she’s that interested in managing a more complicated portfolio.  There’s real value in simplicity if I get older and start to lose my mental faculty, or if I’m unexpectedly dispatched.

Alternate Allocation – Three Fund Portfolio

The three fund portfolio would replace my current stock allocation target with a 60% VTSAX and 40% VTIAX.  The bond portion would stay just the same, and I’ll ignore the little bit of REIT for now.

I just have this idea from books I’ve read that my portfolio is likely better than this simple three fund design.  But is it?  We don’t know what the future will bring, but I have really specific data from the past.  Below I’ll compare my actual 2017 returns with what I would have received under this alternate allocation

Comparison Analysis: Three Fund Portfolio

I keep track of the internal rate of return for my whole portfolio as well as each individual component of the overall strategy.


My Target Portfolio gave me better performance than the alternate Three Fund Portfolio would have in 2017.


Here’s my personal returns for 2017, based on the target allocation described above.

End of Year Allocation Asset Return
67.98% All Stocks 21.49%
38.70%     US Stocks 18.40%
18.85%           US Large Blend 17.40%
5.57%         US Large Value 24.41%
8.81%         US Mid Value 22.15%
5.46%         US Small Value 19.08%
29.28%     All International 26.46%
12.76%         Foreign Large Blend 29.26%
5.36%         European 14.36%
5.60%         Pacific 25.72%
5.57%         Emerging 24.26%
26.89% Bond 2.35%
5.13% REIT 3.94%
 100% TOTAL 18.89%

Let’s replay the year, making all the same contributions on the same dates.  But instead of breaking US Stocks and International Stocks into sub categories, just by the “parent” index in the three fund portfolio – either VTSAX or VTIAX.

Here’s the same table of results compared with the Three Fund Portfolio:

Asset Class My Actual Return Three Fund Return
US Stocks  18.40% 18.65%
International Stocks 26.46% 27.58%
Bonds 2.35% 2.35% (the same)
TOTAL 18.89% 19.23%

Conclusion: Should I Switch to a Three Fund Portfolio?

My hypothesis is rejected:  The three fund portfolio outperformed my own target portfolio for 2017 (19.23% vs 18.89%).

Reasons to Stay With My Current Allocation

Despite the evidence that a simpler portfolio really would have been better in 2017, it is just one year.  I never expected a huge effect, and I certainly wouldn’t expect to “win” every year.  Furthermore, the whole point of an Investment Plan is not to change course on trends and whims.  I think it doesn’t matter as much what your chosen allocation is, it matters if you stick to it through thick and thin.  We have a six month “cooling off” period for any changes in our plan – and the changes have to be in writing first!

We might have more tax loss harvesting opportunities with a variety of funds, rather than just the big three.  Also, I’m employed in a “growth” industry, so I could argue that investing in value stocks helps me to properly diversify my resources.

Reasons to Switch

Investing is simple.  Complicating my plan just to make it more interesting is silly.  I could save a lot of time tracking data and managing different holdings across different accounts.  And it appears I could do that without really losing any returns – the three fund portfolio funds do have lower fees, which ultimately may drive their superior performance.

Finally, I had the idea to explore this topic when I was thinking about my estate plan.  I like managing all the moving parts, but I think those who survive me would prefer a simpler picture.

So I think I’ll switch to a three fund portfolio.


I would sincerely appreciate any arguments for or against; especially any factors I haven’t considered yet.

11 Replies to “Should I Switch to a Three Fund Portfolio”

  1. I always love your analytical ways to make a decision. I used to have a similar complicated setup and I went to 4 fund for my retirement accounts (small cap tilt) and schwab intelligent for my taxable so they can do the TLH.

    More effort rarely pays off in this investing game. You said how having a more complicated mix would provide you 0.05 but it did quite the opposite. I’m now in the less is more camp about tax loss harvesting … let the computer figure it out and I am yet to see my first wash sale penalty.

    1. I agree with you on the less is more. I won’t regret simplifying things. For robo-TLH, I just can’t do it – but it’s because I can’t stand to pay advisor fees to anyone. I really think you just go 100% passive investing at the lowest cost possible and screw anyone trying to take a fee out of your assets!

      Anyway, haven’t done much TLH yet, as most of my savings are tax advantaged. Thanks for your comments, as ever. -M

  2. Yo Mouse, great post. I’d love to hear your thoughts on asset location as well! I have a similar asset allocation (inspired by the Gone Fishing Portfolio), but haven’t given much thought to the location across different vehicles. That’s my next step!

    1. I mentioned on my other comment, but most of my assets are tax advantaged, so I haven’t had too many hard decisions to make on the asset location point. I understand that many people keep bonds in tax advantaged spaces, and international funds in taxable accounts — which fine, that’s efficient. However, you’re also keeping your lowest growing assets in the best space and your fastest growing assets in the worst space so… hmm.

      Probably just like asset allocation, asset location comes down more to 1) have a plan and 2) stick to it. But I need to learn more before I have something substantive to say.

      Thanks for the feedback, cheers -m

    1. I read the article you referenced and appreciate Gasem’s work. I think the framework of having a “safe portfolio” that meets your required expenses and then a more aggressive portfolio for stuff you want (vacations!) but don’t need is a pretty darn good way to look at things.

      Gasem did an analysis by kicking off a hypothetical with a -3 sigma event. That means you retire right at the moment the market has it’s worst year in 300 years. It’s not surprising that a lot of withdrawals rates have trouble handling such a massive event — this truly is a worst case scenario.

      I’d suggest that after such a draw down, subsequent years are more likely to be better than average. When the panic and hysteria fades, folks will want back in to buy all these great companies that are on sale. Furthermore, it’s not clear to me if Gasem is rebalancing his portfolio, or if stocks and bonds are moving independently (I confess I’m a little short on time, so I may have just missed these details). In any case, we’re looking at something like a 90% stock drop in one year — perhaps bond returns are flat? If that’s the case, your 50-30-20 balance turns into 18/11/71. ahh! rebalancing will move a lot of your safe bond money back into stocks, which i think are likely to have a nice bounce in the coming years.

      But I digress. I think if we’re trying to plan retirements around timing the worst year in 300 years for the moment we retire, we might be focusing too much on a hypothetical and ignoring more pressing and realistic problems that can and will destroy a nest egg — Divorce, Addiction, Speculation.

      Thanks for your comment, it got dropped in my comment system’s spam for some reason, so I apologize for the delay. That’s cleaned up now, and I appreciate your ideas 🙂 -M

  3. Did you know the return on the S&P since 2000 is 3.4%/yr? Did you know inflation adjusted the S&P return is 0.9%/yr since 2000? This means if you retired in 1999 with a presumed guaranteed 4%/yr safe withdrawal rate, it’s highly unlikely you will see 30 years of safe withdrawal. My calculation is you run out at 27 years. My point is this is not some 1000 year flood scenario. There are people today living the 1000 year flood scenario. Can you imagine given the present state of affairs a 20% correction in the next couple years? That my friend is a real time 3 sigma scenario 2000, 2008, say 2020 to a 30 year portfolio with a “safe” 4% WR.

    The thrust of my article was to illustrate with more granularity what optimistic withdrawal choices and SORR does to to the longevity of back of the envelop portfolio calculations like 4% x25. Every portfolio calculation to be credible needs to account for both return and risk. The two are linked 100% like every point on a plane has both x and y coordinate. In 2008 a SPY portfolio fell 37% meaning it had to grow 74% to get bask even. A 50/50 bond/spy account fell only 8% and need grow only 16% to realize parity. This means the 50/50 account hits parity and starts compounding again long before SPY gets even. Risk matters. In 2000 it took13 years for SPY to reach it’s 1999 high. 13 years is 43% of a 30 year retirement. That’s a long time to be under water.

    My article also analyzes the risk between a 2 fund portfolio and the 3 fund Bogelhead portfolio. The 2 fund portfolio lives on the efficient frontier and so does not pay for return with too much risk premium. The Bogelhead 3 fund is not efficient, does not live on the efficient frontier and pays too much risk premium for its return. It actually pays 33% more risk for exactly the same return. I then stress the difference between the efficient 2 fund portfolio and the over risked 3 fund portfolio and the results are obvious. The only thing the 3 fund buys you is less safety. I read your analysis regarding international assets. SPY is up 5% this year, EM is down 4.5% and International is down about 1.5%. Over time international has a higher risk and EM has a way higher risk and both lower returns so I don’t see much point in owning them in a simple index portfolio.

    People think all of these funds buy you more diversity, they do not. I call it pile higher and deeper. A portfolio realizes market risk with as few as 20 stocks. 1000 stocks improves the diversity by only 4%. Once you hit market risk you are at market risk, the function is asymptotic and doesn’t improve. In fact owning too many funds can cost you in fee drag.

    I enjoyed the article, written at a high level of analysis.

    1. From June 2000 to June 2018, the S&P 500 annualized return is 3.566%. If you reinvest dividends, it’s 5.531%. Adjusted for inflation, that becomes 3.338%. So — yeah, a little rough if your withdrawal rate is 4%, but then you also picked a starting point right at the top of the longest bull market in history. Pick your starting point in June 2001 instead, and the annualized return adjusted for inflation becomes 4.671% (with dividends reinvested).

      My point on the 1 in a 300 year thing is that a 3 sigma event by definition is a 99.7% event — that’s about 1 in 300. The odds that this year will be the worst year in 300 years are just around there, and I think the following year is not exactly independent of such a calamity.

      But sure, if you retroactively pick the worst moment to retire and then don’t reinvest dividends, your returns have looked pretty slim.

      On your other question, Yes! I can imagine the market pulling back 20% as it’s done before and will do again. That’s more like a 1.5 sigma event and probably happens every ten years or so. It’s a bummer if you’ve got your whole nest egg complete and you’ve just retired. But it’s also a terrific opportunity for those in starting out their earning years. Indeed younger investors should be cheering for a bear market.

      You mention a 100% SPY portfolio losing 37% in 2008 and needing 74% gains to get back to even. Indeed, if you just held your SPY account in June 2008, it would be back to even by January 2011. Continue to hold it through present day, and you would be well compensated for your risk: 118% gain with dividends reinvested (that’s 8.1% per annualized).

      The 50/50 portfolio would not have sunk as low, and would have broken back to even more quickly, but would only have enjoyed an 80% gain (depending a bit on if/how you rebalanced) — about 5.5% per year.

      I’d conclude that if you needed to access the money, you should be in the safer allocation, but if your needs were satisfied, you’d enjoy better returns with the riskier portfolio, yes?

      Also, I 100% agree with you – get that market risk and minimize your costs. I think the fees are low enough on the international index funds that it’s worth it; but I avoided international bonds for just that reason. I don’t think they add much diversity so why pay more? We’re ditching EM/Pacific/European and value stocks to minimize our fees. I’ll read more comparing 2-fund to 3-fund, but with our current administration I might like to hold some interests outside of the US 🙂

      Cheers, and thank you very much for your kind and thoughtful comments! -M

  4. I think you measure the S&P wrong. You measure trough to peak but I think the correct measurement is local peak to local peak to wit the local peak in 2008 was Oct 2007. You didn’t get even till April 2013, 6 years later. My own portfolio which is risked at 2/3 market risk was even in 2011 and 18% ahead in 2013. The local peak prior to 2007 was March 2000. I didn’t cherry pick, that’s just the market data and I lived through both. A man who retired Jan 1, 2000 with a 4% WR has an excellent chance of being hosed before 2030 (inflation adjusted). Again my point is SORR is not necessarily a singular sentinel event. A long period of under performance is just as deadly.

    Your comment on retiring in 2001 is exactly the thought behind my proposal of portfolio insurance:

    Once retired how do you re-sequence the SORR without going back to work? Answer a risk free insurance policy. Closing the portfolio to withdrawal allows re-balancing and reinvestment of dividends into stocks and revitalizes the stock part of the portfolio for a better outcome effectively re-sequencing the SORR. Buy low sell high. The cost is only 2 extra payments of WR so you would retire on 4% x27 instead of 4% x25.

    PS. I don’t have a dog in the hunt, I’ve just done a lot of work on looking at what I think is optimum based on quant modeling.

    1. I just measured from June 2008 to break even. I’m happy to start on any date and end on any date, but I’m not sure how productive it is to assume a start date of “absolute worst moment” for the analysis.

      It’s hard to know what the future will bring, and indeed even the efficient frontier only looks backward. However, the good news is we can exactly test how different portfolios would have fared in history. You give an example of a man who retired on Jan 1, 2000 with a 4% WR. How would that man be doing today? We can find out exactly.

      It really merits its own full analysis and post, but I’ll do a simple one here. Let’s start with $1million and withdraw $40,000 on January 1st every year. We’ll adjust for inflation and reinvest dividends. 100% S&P.

      I found that such a man would have only $294,000 left in his original nest egg. Indeed, it appears the retiree would be in trouble by the year 2030.

      We are in agreement, I think. The question is how such an example should influence current and future behavior. For me, I have some contingency plans so that if I do encounter a bad sequence of returns and just the wrong time, I’ll be more adaptive and able to survive.

      Note also, that I assumed this retiree withdrew $40,000 each year. If instead they looked at their account and withdrew 4%, they’re in substantially better shape. The account balance on January 1, 2018 would then be $865,000; though they would have endured several years of living on less than $40000.

      A riskless insurance plan could include social security as a back stop for basic needs, and in “bad years” you simply live on less.

Share Your Thoughts

This site uses Akismet to reduce spam. Learn how your comment data is processed.