A donor-advised fund (DAF) is an investment account that allows you to take a tax deduction now and give the money to charity later. When you give money to a DAF, you can deduct that money just as you would deduct a charitable contribution. The DAF invests the money tax-free until you are ready to donate it to charity. But DAFs only allow limited investment options. How can we best make use of a DAF to optimize expected investment performance?
- What investment options do DAFs provide?
- Investment strategies
- (Even more) advanced ideas
- Comparing the choices
- Appendix: More ideas
Before we decide how to invest in a DAF, first we must ask: Should we use a DAF at all? Or should we keep our money in a taxable account? Or should we simply donate it all right away? The answer depends on the circumstances. I review some pros and cons of DAFs in Donor-Advised Funds vs. Taxable Accounts for Patient Donors. In this essay, I will assume we’ve already decided to use a DAF.
I only cover DAFs in the United States because that’s what I’m familiar with. Many other countries probably have similar investment vehicles.
I am not a lawyer or an investment advisor and none of this should be taken as legal or investment advice.
What investment options do DAFs provide?
The three most prominent donor-advised fund providers are Vanguard Charitable, Schwab Charitable, and Fidelity Charitable. A fourth provider, Greater Horizons, is worth including for reasons we will discuss below. All four DAF providers charge similar fees. Fees matter, but these four providers have similar enough fees that we can ignore them for now.
The providers offer index funds covering the following markets:
If your account has over $250,000, Fidelity and Schwab (but not Vanguard) allow you to appoint a financial advisor to manage the account. This advisor can invest in things other than the pre-selected funds, as long as they abide by certain restrictions (which are documented here for Schwab and here for Fidelity). Greater Horizons allows you to open an advisor-managed account with no minimum account size, and imposes fewer restrictions (see here).
I believe Greater Horizons is the most likely to allow advisors to follow risky or unusual strategies1. Possibly due to legal restrictions, none of the DAFs allow investors to use leverage via margin, futures, options, or swaps2 (but they do allow leveraged ETFs).
For more what different DAF providers offer, see A Comparison of Donor-Advised Fund Providers.
Let’s review some possible investment strategies. This list includes all the ideas that I personally have considered, but I’m sure there are some good ideas that I don’t know about.
Baseline: global market portfolio
If the efficient market hypothesis is strictly true, then we should invest in the global market portfolio. The investment selections available in a Fidelity, Schwab, or Vanguard DAF don’t include every asset class in the world, but they do include most of the biggest ones. A DAF can get pretty close to the global market portfolio with something like this:
If you use Vanguard, you can include international bonds, so you can get even closer to the global market portfolio:
Many altruistic investors can tolerate more risk than this. We can increase risk by removing bonds and only investing in stocks—something like 50% in a US stock fund and 50% in an international stock fund.
Next, let’s look at a few strategies that might improve on this baseline.
Caveat: We can never know for certain how well an investment will perform in the future. For the rest of the strategies in this essay, we have at least some reason to expect them to outperform the global market portfolio in the long run, but nothing is guaranteed.
Overweighting emerging markets
In theory, the best way to increase expected return is to use leverage: borrow money to invest more than you have. But most DAFs don’t allow leverage. The only way to increase expected return is to tilt toward high-risk, high-return asset classes. We can do this by investing in all stocks and no bonds. Can we go further?
To increase expected return, we can concentrate in the region of the world where equities have the highest expected return. To do so, we need to make an educated guess as to which region that is.
We don’t have any perfect way to determine future expected returns, and we could take many different approaches. Several institutions provide return projections using various methodologies, and they generally agree that emerging market equities currently have higher expected returns than other asset classes (e.g., AQR, Research Affiliates, Vanguard). This higher expected return comes with higher risk. So we can simulate leverage by buying emerging market stocks.
(Of the major DAF providers, only Vanguard offers an emerging markets index fund as a pre-selected option. Investors who want to hold emerging markets, but don’t want a professionally managed account, would need to use Vanguard.)
How much higher of a return should we expect from emerging markets than from global equities? That depends on what methodology we use to estimate future returns. According to AQR, emerging markets have an expected excess return of about one percentage point, while Research Affiliates expects a three percentage point excess return (as of March 2021).
The main reason for this difference is that Research Affiliates adjusts for stock market valuations, while AQR does not. Emerging market stocks look cheaper than stocks in developed markets. We know that cheap countries tend to outperform, sometimes by a lot (Faber, 20123; Radha, 20204). More generally, cheap assets tend to outperform expensive assets both within and across markets (Asness et al., 20135). There’s mixed evidence on what causes this, but it could happen due to a combination of home country bias (Swedroe, 2019; Saphier et al., 2019), common behavioral biases (Lakonishok et al., 19936), and limitations on sophisticated investors that prevent them from correcting this mispricing (Lewellen, 20117; Shleifer & Vishny, 19978). That said, Research Affiliates’ estimates might over-correct for the valuations of different markets, so my own estimate for the expected return of emerging markets would fall somewhere between AQR’s and Research Affiliates’ estimates.
Momentum across asset classes
As discussed in the previous section, cheap assets tend to outperform expensive assets. Value isn’t the only investment factor that systematically predicts performance. There’s also the momentum factor: assets that have outperformed over the past 6-12 months tend to continue to outperform for a few months (Jedageesh & Titman, 20119; Fama & French, 200810; Berger et al., 200911). Similarly, assets that have performed well on an absolute basis tend to continue to perform well (D’Souza et al., 201612; Lempérière et al., 201413; Babu et al., 201914; Faber, 201315))—this is known as absolute momentum or trendfollowing.
A DAF can’t invest in individual high-momentum stocks, but it can invest in whichever of the available funds show the strongest momentum. Gary Antonacci describes a version of this concept in his book, Dual Momentum, and in a paper16; a similar approach was studied in D’Souza et al. (2016)12. (The strategy is called Dual Momentum because it uses both relative and absolute momentum.) Dual Momentum works like this:
- Start with three index funds: a US stock fund, an international stock fund, and a bond fund.
- If both stock funds had negative total return over the prior 12 months, buy the bond fund.
- Otherwise, buy whichever stock fund had higher return over the prior 12 months.
- Rebalance monthly.
In a backtest from 1950 to 2018, Dual Momentum performed much better than a diversified buy-and-hold strategy:
|Buy and Hold
We don’t have to use US stocks + international stocks + bonds as our asset classes. A similar strategy would have performed well on many different combinations of asset classes. For example:
- Faber (2010)17 finds strong historical performance for a momentum strategy that rotates among US stocks + international stocks + REITs + commodities + bonds17.
- Moskowitz & Grinblatt (1999)18 identify a momentum premium across industries.
The momentum premium shows up everywhere5—it is the most consistent and robust market anomaly. For more detailed evidence that momentum-based (as well as valuation-based) asset allocation strategies work, including a discussion on expected future performance, see Blitz & van Vliet (2008)19.
We could run an asset class momentum strategy in a DAF through Vanguard Charitable:
- Consider five funds: US stock, Europe stock, Pacific stock, emerging stock, and US bonds20.
- Out of these five funds, hold whichever two have the highest prior 12-month return.
- But, if fewer than two funds had positive 12-month return, put the remaining money into a money market fund.
- Rebalance monthly.
Or we could run Dual Momentum in Fidelity or Schwab, using three asset classes (US stocks, international stocks, and bonds) instead of five.
(There are lots of variations of asset class momentum: using a different set of funds; using a time horizon other than 12 months, such as 6 or 9 months; rebalancing quarterly instead of monthly; etc. Every variation has pros and cons. The important thing is to capture the momentum premium, and any reasonable implementation will do that.)
What kind of future returns might we expect from asset class momentum?
As a basic approach, let’s look at how well asset class momentum has performed in the past compared to buy-and-hold. This tells us the historical size of the “asset class momentum premium”. But the premium will likely shrink in the future21. If we roughly assume that asset class momentum will work half as well in the future as it has in the past and that the volatility won’t change, then we can expect asset class momentum to earn about a 5-6% net real return with 12% standard deviation22.
Note: Asset class momentum strategies often move in and out of positions many times in a year. DAFs generally don’t want investors to trade excessively. If they believe you are trading too frequently, they might decide to limit your trades. I expect they wouldn’t do that, but it’s a risk worth considering.
Leveraged ETFs (with a managed account)
By default, DAF providers only allow you to buy funds from their pre-selected list. But most DAF providers allow you to appoint a professional investment manager. (Advisor-managed accounts typically have a $250,000 minimum, although Greater Horizons has no minimum.) That manager can invest however they want as long as they follow certain guidelines. They can’t get leverage via margin, futures, or options, but they’re allowed to buy leveraged ETFs.
Leveraged ETFs tend to cost more than other types of leveraged investments. In a previous essay, I estimated that leveraged ETFs on average cost about 2% plus the risk-free rate (although the exact number varies a lot over time). Investment managers typically charge around 1% for small(ish) investors and 0.5% for large investors, so holding leveraged ETFs in a managed account will cost 2.5 to 3 percentage points more than investing in a DAF’s pre-selected funds. Fidelity Charitable allows you to self-manage your DAF if you have over $5 million, so you could cut out the advisor fee.
For truly risk-neutral investors, this is probably worth it. Under most plausible assumptions, leveraged ETFs have a high enough expected return to make up for the cost. For investors with logarithmic utility (which is probably still too aggressive for individual investors, but might be appropriate for altruists), the higher cost of leveraged ETFs probably cancels out their higher expected utility. (For details, see The True Cost of Leveraged ETFs.)
Factor funds (with a managed account)
The value and momentum factors have historically predicted stock performance. (There are some other factors as well, such as size, low-volatility, and quality. These aren’t as well-supported as value and momentum, but they might be worth incorporating.) Some ETFs and mutual funds provide exposure to these factors. With a managed account, your investment advisor could use factor funds to try to outperform the market.
Factor funds (at least the good ones) are much cheaper than leveraged ETFs. To be worth buying, factor funds need to outperform the market after both fund expenses and the advisor’s fee. That’s probably achievable in expectation.
You can run asset class momentum yourself (and skip the advisor’s fee), so it seems less likely that a factor fund portfolio can outperform asset class momentum after costs. But using an advisor means you don’t have to do any work, and the time savings might justify the extra cost.
Concentrated stock picking (with a managed account)
If we’re willing to take heavier bets on value and momentum, we can concentrate our investments in a small number of value/momentum stocks (follow the link for a more detailed explanation). Backtests suggest that the top 10% of small-cap value and momentum stocks earned about double the return of the market before fees (with more than 1x the volatility of the market, but less than 2x). A portfolio of individual stocks would have to pay trading costs plus a management fee. This would probably cost less than a leveraged ETF, but definitely more than an un-managed DAF portfolio.
How good is a concentrated value/momentum strategy? That depends on how well we expect the value and momentum factors to perform in the future. It seems likely that they will continue to work at least to some extent, and the value factor in particular looks attractive relative to history (as of early 2021). Still, it’s difficult to forecast future performance. But even if a concentrated value/momentum strategy has zero risk-adjusted premium over the market, it should still outperform a leveraged ETF by having lower costs.
A related approach: Instead of buying individual value/momentum stocks or a set of diversified factor funds, we could buy a concentrated mutual fund or ETF that looks particularly attractive. For example, as of this writing, the Russian stock market looks cheaper than almost any other market, so we could buy the iShares MSCI Russia ETF (ERUS). Buying ERUS is the same basic strategy as buying an emerging market index fund as discussed previously, but it’s higher risk and has the potential to deliver a higher return. We can’t get as much concentration with an ETF as with a basket of stocks, but this approach has the advantage that we only need to buy one thing.
(Even more) advanced ideas
Create a new DAF
Legally, any registered charity can operate a DAF. (At least that’s my understanding from reading IRS guidelines.) The charity just needs to accept money from donors, invest that money, and allow donors to recommend grants to other charities. But it’s a lot of work to set up all the necessary infrastructure. A sufficiently motivated organization can do it if they believe they’ll get enough assets under management to justify the effort.
Charities have to abide by certain legal restrictions about how they can invest, but the restrictions are not well-specified and I don’t understand them very well. A new DAF might be able to provide better investment options than any of the ones I’ve discussed so far, or it might not, depending on the legal situation.
Launch a new leveraged fund
The leveraged ETFs on the market today have some problems that make them less than ideal for long-term investing. But a new leveraged ETF or mutual fund could fix these problems. In particular:
- Existing leveraged ETFs charge high fees. Most people who buy leveraged ETFs use them for short-term trading, so they don’t care much about the fees. But these ETFs have high profit margins, so a few fund could undercut them while still making a profit.
- Existing leveraged ETFs rebalance daily. A long-term risk-friendly investor doesn’t need their account to perfectly track the leveraged index on a daily basis. A new leveraged ETF could rebalance less frequently (perhaps monthly), which would reduce trading costs.
- Existing leveraged ETFs focus on a single market or sector. This makes each individual fund riskier. It also means an investor who buys multiple leveraged ETFs will need to rebalance their holdings manually. But a new leveraged ETF could invest in the global market portfolio (or something like it). That way, long-term investors could buy a single fund and would never need to rebalance.
Investors with a minimum of around $25 to $50 million could sponsor the creation of a new leveraged fund. Such a fund might be a better way to invest than any of the other options we’ve discussed.
Comparing the choices
So far we’ve looked at seven investment strategies that might perform better than the global market portfolio:
- Buy and hold an emerging markets index fund.
- Run asset class momentum on the pre-selected asset classes.
- With an investment manager, buy and hold leveraged ETFs.
- With an investment manager, buy and hold some factor funds.
- With an investment manager, invest in a concentrated basket of value/momentum stocks or a concentrated ETF.
- Create a new DAF that offers more investment options.
- Sponsor the creation of a new leveraged ETF or mutual fund.
I don’t know what legal restrictions would apply to a new DAF, so I can’t say much about that idea. Creating a new leveraged fund looks like the best option, but isn’t possible for most investors. Existing leveraged ETFs seem not worth buying. That leaves four choices: buying an emerging markets fund, running asset class momentum, buying factor funds, or buying a concentrated basket of stocks. Investors who do not meet the minimum requirement for a managed account (usually $250,000) must pick between the first two choices.
When possible, I would prefer to invest in a concentrated basket of stocks. Historically, they offered substantially higher returns than the market, and I expect this to continue (although probably to a lesser extent). Buying and holding factor funds is less risky, but also has worse expected return. You’ll have a much easier time finding an investment advisor who’s willing to buy factor funds, so it might be a more realistic option. (Buying factor funds is a pretty normal thing that lots of advisors do. Buying a few dozen deep value/momentum stocks is a weird thing that almost nobody does.)
Between the two strategies that do not require a professionally managed account—asset class momentum and an emerging markets index fund—I lean toward asset class momentum, for three reasons:
- Asset class momentum diversifies across multiple asset classes. All else equal, diversification is good.
- According to backtests I ran, if we went back in time to 1990 and bought the asset class that performed best over the next 30 years, we still would have underperformed asset class momentum. Even with perfect foresight, asset class momentum still beat buy-and-hold.23
- According to my (very unreliable) future return projections, asset class momentum has a higher expected return than emerging markets.
I don’t put much credence in the last two points, but they provide at least weak evidence in favor of asset class momentum.
On the other hand, it seems hard to deny that emerging markets currently have a higher expected return than other parts of the world. Asset class momentum seems likely to continue working, but I’m not quite as confident about that. Holding an emerging markets funds has the added advantage that you never have to rebalance, so if you want to do as little work as possible, it might be a better choice.
Appendix: More ideas
Hire an advisor and let them decide how to invest
In general, this is a bad idea. Most investment advisors cannot outperform the market, and they’ll charge you a fee for essentially no added value. You’d be better off using the DAF’s pre-selected funds to build the global market portfolio (or get as close to it as you can).
If your advisor is familiar with the research on value and momentum investing, they’re more likely to invest your money well.
Find an advisor who can generate alpha
It’s not too hard to find an investment advisor who can build you a portfolio of value/momentum factor funds that’s better than the global market portfolio. But it’s much more difficult to find someone who can generate true alpha on top of those factors. A few such people probably exist, but they’re nearly impossible to identify before the fact. In general, I don’t think it’s worth trying, although there might be exceptions.
I spoke to representatives at each donor-advised fund and to a few financial advisors who manage DAFs. Based on those conversations, it appears that Greater Horizons offers the most flexibility. ↩
Unfortunately, the legal requirements are not clear. A DAF provider has a fiduciary duty to make prudent investments, and it’s largely up to courts’ interpretation as to what this means. DAF providers disallow investments that they think might violate fiduciary duty. For more, see the FDIC Trust Examination Manual: Investment Principles, Policies and Products. ↩
Faber (2012). Global Value: Building Trading Models with the 10 Year CAPE. For more detail, see Faber’s book on the same subject, Global Value: How to Spot Bubbles, Avoid Market Crashes, and Earn Big Returns in the Stock Market. ↩
Lewellen (2011). Institutional Investors and the Limits of Arbitrage. ↩
D’Souza,, Srichanachaichok, Wang & Yao (2016). The Enduring Effect of Time-Series Momentum on Stock Returns Over Nearly 100-Years. ↩ ↩2
Faber (2013). A Quantitative Approach to Tactical Asset Allocation. ↩
Blitz & van Vliet (2008). Global Tactical Cross-Asset Allocation: Applying Value and Momentum Across Asset Classes. ↩
This list does not include the international stock fund because that fund is simply the sum of the Europe, Pacific, and emerging stock funds. Unlike Antonacci’s Dual Momentum, I treated bonds as a risky asset. I believe Antonacci’s backtest makes bonds look better than they are because bond yields consistently decreased for most of the tested period. It makes more sense to use a money market fund as the risk-free asset. ↩
We should expect the asset class momentum premium to shrink for two main reasons:
- Bond yields are lower than they were in the backtest period. That reduces how much return an asset class momentum strategy can earn when it’s not holding stocks.
- Momentum strategies have become more popular, which could cause them to perform worse going forward.
The first point seems hard to dispute. Regarding the second point, as a simple rule of thumb, we could assume the premium will decline by half24. Relevant research:
- McLean & Pontiff (2016)25 found that published anomalies historically shrunk by 58% after publication, of which 26% could be attributed to data mining, and 32% to markets becoming more efficient.
- Jensen, Kelly & Pedersen (2021)26 expanded McLean & Pontiff’s research and found about a 1/3 decline post-publication, all of which could be attributed to data mining.
- Blitz (2017)27 found that ETFs do not harvest the momentum premium28.
Asset class momentum appears more likely than stock momentum to survive increasing popularity because (1) it trades larger, more liquid positions, and (2) most momentum investors use stock momentum, not asset class momentum. So arguably we should make a more optimistic assumption. ↩
I came to this expected return figure as follows:
- As a baseline, assume Dual Momentum will perform as well as the average of the US and international stock markets. If we take the median of AQR’s and Research Affiliates’ projections, we get an expected return of 3.4%.
- Based on backtests, the historical Dual Momentum premium was about 6%. Halve this to get 3% and add to the baseline performance for a total expected return of 6.4%.
- Subtract 1% for transaction costs to get 5.4%.
I backtested two different pairs of asset classes: (US stocks, global ex-US stocks) and (US stocks, Japan stocks). I specifically chose Japan because Japan had the weakest performance of any major stock market over the sample period. If asset class momentum over (US, Japan) outperformed US stocks alone, that would give the strongest possible evidence in favor of asset class momentum that my data set could produce. And indeed, it turned out that asset class momentum did (slightly) outperform. I don’t consider this strong evidence in an absolute sense, but it’s the best I can do given the data I have access to (I used the Ken French Data Library). ↩
In How Can a Strategy Everyone Knows About Still Work? , Cliff Asness suggests making this “conservative” assumption. ↩
McLean & Pontiff (2016). Does Academic Research Destroy Stock Return Predictability? ↩
Blitz (2017). Are Exchange-Traded Funds Harvesting Factor Premiums? ↩
We might be interested in how hedge funds behave rather than ETFs, because hedge funds probably represent more ‘savvy’ investors. I am not aware of any research on whether hedge funds in aggregate exploit the momentum premium. But another 2017 study by Blitz, Are Hedge Funds on the Other Side of the Low-Volatility Trade?, found that hedge funds systematically reverse load on the low volatility premium (that is, they take the losing side of the trade), which weakly suggests they may do the same for momentum. ↩