[Editor's Note: This is one of my columns written for MDMag.com (Formerly Physicians Money Digest) and grew out of a conversation I had with Rick Ferri at Bogleheads 12 about adding new asset classes to a portfolio. I hope you find it helpful.]
The benefits of having several different asset classes in your portfolio are well-known. There are dozens of increasingly exotic asset classes available to invest in, including frontier market stocks, U.S. small growth stocks, Japanese real estate, and British inflation-indexed bonds. An investor need not invest in all of them in order to be successful in reaching his financial goals.
The point of holding multiple asset classes is to boost returns and decrease risk through diversification. Since neither you, nor I, can predict the future, you want a portfolio that is going to do acceptably well no matter what happens in the world.
Consider these guidelines when deciding whether or not to add a new asset class to your portfolio.
7 Guidelines for Adding New Asset Classes to Your Portfolio
#1 Number of Asset Classes
If your portfolio currently contains only one or two asset classes, you will almost surely benefit from adding another one.
Although the benefit of adding an additional asset class to the portfolio goes down with each new asset class, the benefits can be dramatic initially. As you add more classes, you are weighing the additional diversification against the added complexity and expenses inherent in a more complicated portfolio.
An investor with a tiny portfolio relative to what he will eventually need to retire can start out with a single asset class, since the effect of additional savings will dwarf the effect of the investment returns of the portfolio. As the portfolio grows, he will want to consider additional asset classes to provide diversification.
As a general rule for a portfolio of reasonable size, I would consider a bare minimum of three different asset classes. Seven asset classes provide a great balance between diversification and complexity. There is very little benefit to additional asset classes once you get to 10. Rather than adding an additional class to an already complex portfolio, perhaps you should consider replacing one that you currently have instead.
#2 Avoid Performance Chasing
The addition of new asset classes to a portfolio is often used as an excuse by the investor to engage in the harmful practice of performance chasing. Investors consciously and subconsciously project the recent past into the future, despite the well-known fact that past performance is no indication of future returns. As such, it was popular to add small value stocks and REITs to portfolios in 2003-2007 and to add gold and long-term treasuries to portfolios in 2010-2012.
These days, after years of outstanding stock performance, investors are talking about a “100% equity” portfolio again. While adding any of these asset classes to your portfolio may make sense for the long run, be sure to carefully examine your motivation to ensure you’re not just chasing performance. If you’re convinced an asset class belongs in your portfolio, consider adding it after a period of poor performance, rather than when it is the “hot” asset class.
#3 Low Correlation
When adding an asset class you want to make sure it is fundamentally different from what you already own.
Bonds are loans to companies or governments. When you own a stock, you own part of a company and share in its profits. REITs invest primarily in real estate and have a different tax structure than a more typical stock. It is easy to argue these asset classes are fundamentally different. Likewise, long-term Treasury bonds have different risk characteristics than short-term corporate bonds. You’re looking for asset classes with relatively low correlation with one another.
On the other hand, if you have a portfolio consisting of large growth stocks and small value stocks, adding some large value stocks (which have relatively high correlation with the other asset classes you hold) probably isn’t going to get you the bang for the buck you would get from adding some bonds or real estate (which have a much lower correlation with the stocks you already hold).
Historical correlations are relatively easy to look up. Correlations vary over time, but you’re looking for an asset class that is non-correlated with your current asset classes as much as possible.
#4 Positive Real Returns
Just about anything and everything — including stocks, bonds, precious metals, art, timber, whole life insurance, and horse manure — can be considered an asset class. Low correlation with the rest of your portfolio is important; however, it is also important to have asset classes that are expected to provide a positive, after-inflation return. You’re investing, not just collecting. If too much of your portfolio is in asset classes with a low expected return, your entire portfolio may not have a return sufficient to meet your goals. Gold is a classic example. If you had an ounce of gold 500 years ago, you could use it to buy a man’s suit. Today that ounce of gold still just buys a man’s suit. Given our current historically low interest rates, many fixed income asset classes currently don’t have expected returns higher than inflation. While there is some argument for holding an asset class or two without an expected positive real return in your portfolio simply for the overall diversification effect on the portfolio, you certainly want the vast majority of the portfolio to beat inflation.
#5 Accessibility
For many years, an investor was only able to invest in a few asset classes because others simply weren’t accessible in a way that allowed an investor to be diversified within the asset class at a reasonable cost. If there is no mutual fund or ETF that allows an investor to easily purchase dozens or hundreds of stocks, bonds, or properties at a reasonable price, the asset class really isn’t particularly accessible.
In recent years, dozens of previously inaccessible asset classes have become accessible thanks to the explosion of the index fund and ETF markets, providing investors many new options.
#6 Liquidity
When adding an asset class to a portfolio, an investor ought to consider her liquidity needs. While a typical investor doesn’t need instantaneous liquidity for his entire portfolio, it is important to be able to have a reasonable amount of liquidity in the portfolio for unforeseen personal needs, investment opportunities, and portfolio rebalancing.
If most of your portfolio is tied up in individual real estate holdings, hedge funds, and other private investments, you should lean toward an asset class that can be liquidated any time the markets are open, such as publicly traded stocks and bonds.
#7 Tax Implications
An investor who has very little tax-protected “space” in his portfolio (such as IRAs and 401(k)s) may find that the diversification benefits of a particularly tax-inefficient asset class such as REITs or TIPS may not be worth the additional tax cost. He may instead prefer to invest in tax-efficient stocks, municipal bonds, and individual rental properties.
Likewise, an investor whose portfolio is primarily tax-protected can consider high turnover stock asset classes (like microcaps) and particularly tax-inefficient asset classes such as Peer to Peer Loans. Even within a taxable account, investors with high taxable income are much more likely to benefit from adding an asset class such as municipal bonds than an investor with a lower income.
While a wise investor rarely changes his investment plan, there may come a time when the inclusion of a newly-investable asset class in the portfolio is prudent. Use these guidelines to help you decide whether the benefits of adding an asset class outweigh the downsides.
What other factors help you decide whether or not to add a new asset class to your portfolio? Comment below!
I agree that adding asset classes can improve results and limit risk at the same time based on the efficient frontier concept. Specifically I am curious about how you justify 7-10 asset classes over the 3 fund portfolio advocated by Taylor Larimore? Do you have actual data that supports a slice and dice portfolio? Those three funds have extremely large diversity built in and are the basis of the Target Retirement and similar funds. VTSMX probably includes REITs, gold companies, etc on its own.
If there is convincing data if am all for it, but absent that I much prefer the simplicity of the basic 3 fund portfolio. I also understand that you enjoy the extra level of attention and work associated with the slice and dice method so I’m not at all trying to suggest it is a bad idea.
The actual data supporting a slice and dice approach over a total market approach is primarily the historically higher returns from small and value stocks. I assure you that data is as robust as investing data ever gets. This is generally thought to be a risk story. So you’re taking on more risk in hopes of a higher return that has historically been there. Will taking that risk be rewarded in the future? Probably, but nothing is for sure. You may very well be better off taking a total market approach and keeping costs as low as possible (slice and dice costs slightly more too.)
Of course VTSMX includes REITs and mining companies and small stocks and value stocks. The question is whether it is worthwhile to overweight those stocks with higher expected returns. I think it is worth the additional complexity. Some do not. As Taylor says, there are many roads to Dublin. It is unlikely that this decision will make the difference between your having financial security and not having financial security. Compared to your savings rate, what you pay advisers, and your tax management, this is a relatively small decision.
What you really need to ask yourself is “If a three fund portfolio is better than a two fund portfolio, why wouldn’t a four fund portfolio be better than a three fund portfolio?” It usually is, up until the point where complexity and cost overwhelm the additional diversification benefit.
Thanks. I think your point on the slightly better return versus the slightly higher cost of a slice and dice, combined with my high savings rate, is what keeps me in the simplistic mode.
Here’s a good example of why I slice and dice. Take a look at Madsinger’s monthly posts on Bogleheads where he compares the returns of various portfolios over the years.
http://www.bogleheads.org/forum/viewtopic.php?f=10&t=127523&newpost=1876718
Annualized returns since 1999
3 Fund Portfolio: 5.71% per year
Slice and Dice: 7.51% per year
Coffeehouse Portfolio (another type of slice and dice): 7.18% per year
Sheltered Sam (another type of slice and dice): 7.47% per year
Now some of this is due to the “lost decade” in US large caps effect and it is only 14 years of data, but even if you only allow for a slice and dice portfolio to outperform by 1% per year, over 30 years and assuming a savings of $50K per year and a return of 8% vs 7%, we’re still talking about a difference of over a million bucks in retirement. Small increments of additional return matter a great deal over long periods of time.
Well that is a bit more compelling than I expected. I will have to think on this some more.
If you were to move from a 3 fund portfolio to a more slice and dice portfolio, which asset classes would you add first?
That’s a tough question, and I don’t think there is a right answer. I think Small Value is a good option. REITs are also a nice option. International small is also reasonable. TIPS or international bonds on the fixed income side. A case could even be made for precious metals, especially these days after a rather large drop in their value (although that doesn’t mean they’ll go up next.)
I think the very best diversifier to add to a portfolio consisting only of US large cap stocks, international large cap stocks, and US bonds would probably be real estate. REITs are the easiest way to get that, but also provide the highest correlation to the rest of the portfolio (especially the US stocks). An individual investment property has great diversifying potential, but has its own hassles and risks. Privately syndicated real estate offerings can also make sense, but there are additional risks and costs involved.
Why do you believe you can predict asset correlation going forward?
Obviously correlation varies over time and there is no way to predict it going forward. The best example is probably Fall 2008 when all risky asset classes fell at once. The only diversification that mattered was that provided by treasuries and other very safe asset classes.
I’m sorry if careless wording left you with the impression that I believe future correlation can be predicted any better than future returns. That said, what do you use to determine any type of expected correlation in the future aside from past correlation?
The point you often make, that future market direction is unpredictable, is a very good one. Diversification is also a very good point.
How is picking and choosing asset “diversification” slices any different (mathematically) than betting on specific stocks, sectors, or bonds.
This road just leads to all kinds of complicated hedging strategies, etc. Seems much smarter to just buy everything and forget about trying to infinitesimally slice and dice, weight, tilt, etc. Will you miss out on that huge rally in some micro slice, sure, but it wouldn’t be that big of a percentage of your portfolio to matter one way or the other, and all the additional overhead of buying, selling, rebalancing, increased fees, doing due diligence on all of these slices, etc…I can’t see how that’s worth it – vs just using that time to earn more money or enjoy life more.
You make a strong counterargument to slicing and dicing. It is definitely a controversial topic.
The benefit of additional assets classes depends on how they interact with the larger portfolio. The asset-asset interaction (referred to as correlation) of some classes is nearly identical. For example, some on the Boglehead forums say Small Value and REITs tend to move together due to overlapping components, so there is little point in having both.
Can you list some asset classes which play off each other nicely? Thanks.
I disagree that small value and REIT are the same thing. Take a look at this year’s returns. SV up several percent more than a great year for the S&P 500 and REITs barely making any money at all. They clearly don’t have that high of a correlation this year.
Are MLP’s considered a “good” adjunct diversification asset to a physician investor?
An MLP is simply a structure, like a mutual fund or a variable annuity. The asset is what is inside it. Most MLPs involve the production, transport, or refining of fossil fuels, so the asset is generally the company working with the commodity. Returns have been quite attractive in the past. The tax consequences of the structure are interesting, there are advantages but it introduces a lot of complexity. I believe there are now ETFs that own MLPs. This simplifies the structure, but also eliminates some of the tax advantages. I don’t see a problem with a physician using “MLPs” as one of his asset classes. I don’t own any. I don’t think it is mandatory to own them to achieve financial success.
I have been doing a lot of reading and thinking about this over the last couple of days and, for me, I think the likelihood of making an error is greater than the possible reward of a point or two of possible extra earnings. As it is I should end up with more than I need anyway so chasing performance and possibly making mistakes will only hurt me. I enjoy saving and investing as well as learning about it, but I am not quite to your level of enthusiasm ;).
Going back to some Boglehead sayings like “if your investment plan can’t be explained in one or two sentences, it is too complicated” and Rick Ferri even agreeing in his own thread based on his recent article on the addition of international stocks that the benefits can be statistically negligible I am sticking with my more simple method. With 4 tax advantaged accounts (me – TSP, IRA; wife – 401(k), IRA) and taxable accounts it would be more complicated to balance it all out for 7-10 asset classes and funds. The way I have it now is Lifecycle in my TSP, 3 fund of Admiral shares and Investor shares that matches Target Retirement in my IRA (to strive for slightly lower ER at the cost of manual rebalancing), the closest to a TR in wife’s 401(k) and TR in wife’s IRA. Once my wife drops to part time next summer we will roll her 401(k) into the TR in her IRA. The taxable account will morph into a 3 fund with a tax-exempt bond fund as it gets built up. I sure wish they had a tax exempt bond LifeStrategy or Admiral TR funds …
I do agree that diversification is a good thing in any investment portfolio. That being said, I believe one can do this with indexed funds and not have up to ten different funds.
Vanguard Total Stock Market will cover small, mid and large cap US equity. Vanguard Total Bond Market covers a wide variety of corporate, government bonds of various maturities. Add a splash of a Total International Equity fund and you just about have the gamut covered. You could argue a slice or a REIT index and/or metals funds and maybe international bond fund.
Investors, lets don’t make this more difficult than it needs to be. Invest in a few diversified US and International Equity funds, a Total Bond Market Fund. Re-balance annually to the percentages that allow you to sleep at night without having to over indulging in alcohol or lunesta.