Donor-Advised Funds vs. Taxable Accounts for Patient Donors

By MichaelDickens @ 2020-10-19T20:38 (+44)

Update 2021-01-15: While I still believe I identified the most relevant factors for comparing donor-advised funds and taxable accounts, I believe my expected utility calculator has significant flaws, and it should probably not be used.

Confidence: Likely

A donor-advised fund (DAF) is an investment account that allows donors to take a tax deduction now and give the money to charity later. When you put money into a DAF, you can deduct it just as you would deduct charitable contributions. Then you can direct the DAF on how to invest the money, and choose to donate it whenever you want.

If you want to invest to give later, DAFs have some clear advantages, plus some limitations. Is it better to use a DAF, or to keep your money in an ordinary (taxable) investment account?

According to the assumptions made in this essay:

Disclaimer: I am not an investment advisor or a tax advisor. Nothing in this post should be construed as investment advice or tax advice. This content is for informational purposes only. Please do your own research or seek professional advice and otherwise take reasonable precautions before making any significant investment or tax decisions.

How do DAFs work?

(This article mainly covers investing in the United States, because that's what I'm familiar with. I don't know if people in other countries have access to anything like DAFs.)

Investors can create a donor-advised fund through a DAF provider. The three most notable DAF providers are Fidelity, Vanguard, and Schwab. These providers allow you to choose how to invest your DAF from a set of several pre-selected mutual funds. The exact funds vary between providers, but all three of them offer at least a US stock index fund, an international stock index fund, and a US bond index fund.

These providers also charge an annual fee: 0.6% for accounts with less than $500,000, and lower fees for larger accounts (the exact fee depends on which provider you use).

How do DAFs compare to ordinary investment accounts? To answer this question, let's look at three factors:

  1. Fixed costs (taxes and fees)
  2. Risk of complete loss of capital
  3. Investment performance

It's easy to make comparisons within each of these factors. For example, whichever type of account has lower fixed costs is better along that dimension. But it's not immediately obvious how to weigh these factors against each other. So I'll start by looking at the three factors individually and see whether DAFs or taxable accounts come out ahead in each case. Then I'll examine how to combine the factors, and use this to come up with an answer.

Sometimes, investors have more choices than just a DAF or a taxable account (for example, investing through a foundation). But to keep things simple, I'll just compare the two types of accounts.

Fixed costs (taxes and fees)

Ordinary investment accounts and DAFs each have a different kind of fixed cost. Ordinary accounts must pay taxes, and DAFs need to pay administrative fees.

Ordinary investors have to pay capital gains taxes any time they sell an appreciated asset. To some extent, we can avoid incurring capital gains by investing in diversified ETFs and not touching them until we want to donate. Then, instead of selling and donating cash, we can donate the ETFs directly.

But most stock ETFs make regular dividend payouts, and we can't avoid paying taxes on those. Usually[1], dividends are taxed as capital gains. Most people have to pay a 15% tax on capital gains, but investors in the top tax bracket have to pay 20%. Plus, some states charge more on top of that (California taxes capital gains at the same rate as income).

Bonds are taxed at the higher income tax rates, with the exception of municipal bonds, which don't require investors to pay any income tax. Investors who want to avoid taxes can hold municipal bonds. And in fact, altruistic investors might sometimes prefer not to hold any bonds at all, as a way of increasing risk and expected return. Either way, we care more about taxes on stock dividends than on bonds.

(For more on the tax rates for different asset classes, see Tax-Efficient Fund Placement on Bogleheads.)

As of October 2020, a global stock market ETF has a dividend yield of about 2%, which is close to the historical average[2]. An investor with an average income who lives in a state with no capital gains tax must pay a 15% tax on this 2%, which means they lose 0.3% of their portfolio to taxes each year. A wealthy investor in California has to pay 33.3% capital gains tax, which comes out to 0.67% of their portfolio.

Compare that to a donor-advised fund's administrative fees. Anyone with less than $500,000 in their DAF has to pay a 0.60% fee. Bigger DAFs pay lower fees. For example, Vanguard charges 0.30% on amounts between $500,000 and $1 million, and 0.13% on any amount above $1 million.

In short, smallish donors (with less than about $1 million) will pay more in DAF fees than they would in taxes. But for wealthier donors both pay more in taxes and pay less in fees, so a DAF comes out ahead on this comparison.

Taxes on big sales

Another consideration that's relevant to some people: you might have a bunch of appreciated assets that you want to exchange for something else. If you sell them and buy diversified ETFs within a taxable account, you'll have to pay taxes on the gains when you sell. But if you put the money in a donor-advised fund, you probably won't owe any taxes. By giving the appreciated assets to a DAF, you avoid a big tax bill.

This sounds appealing, and does matter under some circumstances. But for a sufficiently patient investor, it doesn't matter. As an illustration: suppose that if you sell and rebalance in a taxable account, you'll have to pay 33% tax, but then you expect your money to compound slightly faster than it would in a DAF—let's say one percentage point faster. If you're willing to wait 41 years, the 1% extra compounding will eventually make up for the initial tax loss.

Risk of complete loss of capital

Any investment account faces some risk that the money ultimately doesn't get spent the way you want. Some types of risks affect both DAFs and taxable accounts. For example, if the financial system collapses, both a taxable account and a donor-advised fund will become worthless. But such risks don't tell us anything about whether to prefer a DAF or a taxable account. So we should focus on risks that affect only one of the two types of accounts.

Each type of account faces its own unique risk. A DAF has a risk of expropriation, and a taxable account risks value drift.

DAFs face higher expropriation risks than ordinary investment accounts for two main reasons. First, DAFs have no legal obligation to follow your grant recommendations. Courts have ruled that donor-advised funds may spend the money however they want, without regard to donors' preferences. It seems unlikely that the large, reputable DAFs like Vanguard, Fidelity, and Schwab would ever ignore donor recommendations, but it could happen.

Second, DAFs can only operate the way they do because the government allows them to. They legally qualify as charities, so donors who give to DAFs can deduct their contributions. And unlike foundations, DAFs are not required to disburse 5% of their funds each year. The government might decide to revoke DAFs' favorable tax status, or might impose limitations on them (perhaps in response to bad press, e.g., 1, 2). That said, such legal changes probably wouldn't destroy the value of all money held in DAFs, so we should be more concerned about the risk that the DAF provider refuses to abide by our recommendations.

While DAFs face higher expropriation risk, taxable accounts are exposed to a greater risk of value drift. If I invest my money in an unrestricted account, I might be tempted to withdraw the money and spend it on myself. If I put the money in a DAF, I have no choice but to give it to charity (although there's still some risk that I'll give it to a "warm fuzzy" charity instead of a highly effective one).

How big of a concern is value drift? Ben Todd wrote an EA Forum post cataloging various attempts to estimate the rate of value drift among effective altruists. We don't have great data, but based on what we know, it seems reasonable to assume about a 5% value drift rate among EAs in general, and a rate of 2-3% among the most committed EAs (which probably includes most of the people reading this).

(Strictly speaking, value drift isn't a loss of capital because you still have the money. But if the money isn't directed toward an altruistic end, then the money is as good as lost for altruistic purposes.)

How about DAF expropriation? According to a Metaculus prediction, DAF providers have about a 0.2% annual probability of expropriating funds, but this estimate probably isn't very reliable. Sandberg (n.d.)[3] estimated that nations face about a 0.5% annual probability of failing, and it seems implausible that DAFs are more than twice as stable as nations. We might roughly estimate the annual probability of DAF expropriation at 1%. An annual probability of 2-3% (on par with the rate of value drift) seems too high, although not totally implausible.

Overall, DAFs appear to have a lower probability of complete loss of capital.

Investment performance

Unless they have a good reason not to, investors should want to diversify their portfolios as much as possible. A fully diversified portfolio includes every asset class in the world. But for our purposes, it probably makes sense to just consider investing in stocks and bonds. Other types of assets are harder to buy, and it's not obvious when buying them is a good idea. Rather than wade into a discussion of which asset classes to include, let's stick with stocks and bonds.

Vanguard, Schwab, and Fidelity DAFs all have the ability to invest in the total world stock market and in US bonds. Naturally, an unrestricted account can also invest in these asset classes, such as by buying ETFs like VT and BND. So if we want to hold the global market, DAFs and taxable accounts are equally suitable. (But if we wanted to hold some other asset class like gold, we probably couldn't do that with a DAF.[4])

Altruists might prefer to invest with leverage. DAFs normally can't use leverage, but taxable accounts can. This is a clear advantage of taxable accounts.

How much leverage should we use? Given altruists' risk tolerance, it might be reasonable for the overall altruistic portfolio to use something like 2:1 to 3:1 leverage (although it's hard to say with confidence, this is just a rough guess). If an investor believes the altruistic portfolio (excluding their own money) already has the correct amount of leverage, then they should use the same amount. If they believe the altruistic portfolio has either too much or too little leverage, then they should compensate by doing the opposite. (I'm inclined to believe most altruists don't use enough.)

Combining the factors

So far, we've talked about three factors: fixed costs (i.e., taxes and fees), risk of complete loss of capital, and investment performance. Now we want to combine them to decide whether we prefer a taxable account or a DAF. We need to figure out how each of the three factors impacts the expected value of an investment portfolio. To do that, we must answer two questions:

  1. How bad is (say) a 2% chance of losing all our money, compared to a 100% chance of losing 2%?
  2. If our investments go up, presumably that means other altruists' investments also went up. How much does that matter?

Let's start with the assumption that we're a small donor. We don't have enough money to substantially impact a cause on our own. That means our last dollar spent does about as much good as our first dollar. So the answer to the first question is easy: a 2% chance of losing everything is approximately as bad as a 100% chance of losing 2%. (For a self-interested investor, or for the altruistic community as a whole, this definitely isn't true—a 2% chance of bankruptcy is much worse than a guaranteed 2% loss.)

Next, how do market movements impact the value of our portfolio? When we gain or lose money, so do most other altruistic investors. When the market goes up, other altruists become wealthier, so our dollars become less valuable. And the opposite is true when the market goes down. There's no simple answer to how much this matters, but we can figure it out with some math, which I cover in the Appendix.

For the reader's convenience, I created a DAF vs. taxable calculator that finds the expected utility of a DAF vs. a taxable account based on the factors discussed in this essay, and allows you to input your own estimates.

Conclusion

Using the calculator I created, we can draw some tentative conclusions about whether to use a DAF or a taxable account.

If I don't believe altruists should use leverage, or if I don't have the ability to use leverage in a taxable account, then a DAF seems better for me. The biggest consideration is value drift. A 2-3% chance per year of value drift probably outweighs DAF fees and the risk that a DAF is expropriated.

If instead I do believe altruists should use substantial leverage (on the order of 2:1 or higher), then the ability to get leverage probably overwhelms all the other factors, and I should keep my money in a taxable account.

Advanced topic: Uncorrelated investments

In a previous essay, I observed that altruistic investors should prefer to minimize their correlation with other altruists' portfolios. But this seems hard to do. I explored a few possibilities, and found that managed futures looked promising as an uncommon investment with low correlation to stocks and bonds.

DAFs generally can't hold alternative investments like managed futures.[4:1] Investors can only hold something like that in a taxable account. Low-correlation investments are so valuable that this probably overwhelms all other considerations (that's assuming managed futures do in fact provide uncorrelated positive return, or that we can find some other investment that does).

Other considerations

There are some other things worth considering that I haven't mentioned yet, including:

In addition, I could be missing other important considerations that would change the result of the analysis.

Appendix: How to calculate expected utility

Consider an altruistic investor Alice. Assume Alice's investments are perfectly correlated with the overall altruistic portfolio. Further assume that her favored cause exhibits constant relative risk aversion. Marginal utility with respect to total altruistic spending is given by

where is the coefficient of relative risk aversion.

Small donors have nearly linear utility of spending. Therefore, the expected utility of Alice's portfolio approximately equals the size of her portfolio multiplied by the marginal utility of the cause (as defined above).

Let be the proportion of Alice's portfolio that she invests in risky assets ( means she uses leverage). For a particular market return , gross utility after one year (ignoring all factors other than market performance) equals

The factor is a constant with respect to Alice's portfolio, so we can ignore it.

Next, we need to account for the other factors considered in this essay. Alice's portfolio must pay some fixed cost, determined either by fees on a DAF or taxes on dividends, depending on the type of account. (For simplicity, I'm assuming the market has a constant dividend yield. This isn't quite true, but it is true that dividends tend to fluctuate much less than market prices do, so it's a reasonable assumption.) So the size of her portfolio after one year is not , but .

Then consider the risk of ruin, due to either expropriation or value drift. In the event of a complete loss of capital, Alice's portfolio can produce zero utility. Therefore, we should discount the expected utility of her portfolio by the probability of complete loss of capital.

Finally, we need to convert the utility of a particular market return into an expected utility over all possible market returns. Assume the market follows a log-normal distribution parameterized by and . The expected value of a log-normal distribution is . Therefore, the expected value of is given by

Accounting for fixed costs and risk of ruin gives a final expected utility of


  1. Qualified dividends are taxed as capital gains, while unqualified dividends are taxed as income. Dividends usually count as qualified, but some types of companies such as real estate investment trusts (REITs) pay out unqualified dividends. ↩︎

  2. Over time, companies have begun paying out lower dividends and replacing them with buybacks, which are more tax-efficient. If companies eventually only return capital to shareholders via buybacks, then investors won't have to pay any taxes at all as long as they never sell any shares. But it seems plausible that at that point, the government will change the rules around buybacks to make them taxable somehow. ↩︎

  3. Sandberg (n.d.). Everything is transitory, for sufficiently large values of "transitory." ↩︎

  4. DAFs may typically only invest in a pre-selected set of mutual funds. Schwab and Fidelity allow DAF owners with over $250,000 to nominate an investment advisor to manage their DAF. But (1) they explicitly forbid using leverage, and (2) they require the portfolio to track an established benchmark, which means you probably can't put a significant fraction of the portfolio into assets other than stocks and bonds. That said, I have not seriously looked into this, so there might be a way to get more flexible investments through a DAF. ↩︎ ↩︎


Jonas Vollmer @ 2020-10-23T17:45 (+3)

Thanks, very helpful analysis! Any thoughts on how much value it would create if someone set up a DAF for EAs that makes leveraged investments? (This is on my longlist of things that EA Funds could try to do.)

MichaelDickens @ 2020-10-23T21:22 (+3)

We can estimate how valuable that would be by comparing the certainty-equivalent interest rates (I talked about this here).

Some quick analysis using leverage.py:

Historical long-run global equities have returned about 5% with a standard deviation of about 16% (source). Let's use that as a rough forward-looking estimate.

  • With a relative risk aversion (RRA) coefficient of 1 (= logarithmic utility), the certainty-equivalent interest rate of an un-leveraged portfolio is 5% (logarithmic utility doesn't care about standard deviation as long as geometric return is held constant). With optimal leverage (2.45:1), the certainty-equivalent rate is 7.7%. That means the ability to get leverage is as good as a guaranteed 2.7% extra return (= 7.7% - 5.0%).
  • With RRA=1.5, optimal leverage = 1.63:1, and the excess certainty-equivalent rate is 0.8%.
  • With RRA=2, optimal leverage = 1.22:1, and the excess certainty-equivalent rate = 0.14%.

I think altruistic RRA is probably somewhere around 1 to 1.5, so under these assumptions, the ability to use leverage is roughly as good as a guaranteed 1-3% extra return.

(FWIW I think you can also get better return by tilting toward the value and momentum factors, so if you're willing to do that, that makes the ability to invest flexibly look relatively more important.)

Jonas Vollmer @ 2020-10-25T12:25 (+2)

(I don't understand how you arrived at 2.45:1 optimal leverage with log utility. I get 5%/(.16^2*1)= 1.95, and the Samuelson formula in leverage.py seems to be the same. Same for the other values.)

MichaelDickens @ 2020-10-26T17:12 (+8)

5% is the geometric mean return, the Samuelson share formula uses the arithmetic mean on the numerator (see here. So the correct formula is (5% + 0.16^2/2)/(0.16^2 * 1) = 2.45.

Jonas Vollmer @ 2020-10-27T09:42 (+2)

Oh, thanks, I was under the mistaken impression that the Samuelson share formula used the geometric mean!

Jonas Vollmer @ 2020-10-25T12:20 (+2)

Thanks! The above calculation compares an un-leveraged portfolio to a leveraged one, but at least under log utility and assuming a low risk of value drift, the relevant comparison is probably between a leveraged (tax-free) DAF and a leveraged taxable account? Presumably, that would be lower than 2.7%.

Also, do you happen to know how effortful and feasible tax loss harvesting might be for leveraged portfolios in taxable accounts?

MichaelDickens @ 2020-10-26T17:17 (+4)

Yeah, because adding leverage will increase taxes on dividends. My calculator correctly accounts for this, but I didn't account for it in my previous comment. But it doesn't lower the certainty-equivalent rate by much.

Also, do you happen to know how effortful and feasible tax loss harvesting might be for leveraged portfolios in taxable accounts?

It shouldn't be too hard, but I don't think you'd get much benefit from it. I'm not sure though, I'm not too familiar with the mechanics of tax loss harvesting.

Jonas Vollmer @ 2020-10-27T10:39 (+2)

Thanks, very helpful.

If we set this up well, we might get $100 million in investments, and the value added would be ~1% excess certainty equivalent rate, i.e., a certainty equivalent of $1 million per year.

If setting up such a DAF takes a year of labor, maintaining it takes 0.25 FTE, and labor has an opportunity cost of $3 million per year, it would take 3/(1-3*0.25) = 12 years to break even (with a plausible range from two years to 'never').

Over a period of ten years, it would return around 10/(1+10*0.25) = $3 million per person-year (with a plausible range from $500k to $5 million).

That seems pretty good, but perhaps slightly less valuable than other things EA Funds could be doing. 

I'd be keen to hear if you think this seems like a reasonable overall takeaway.

MichaelDickens @ 2020-10-27T16:21 (+2)

labor has an opportunity cost of $3 million per year

This seems really high. You could hire an experienced investment manager for a lot less than that. But the general structure of your analysis seems sound.

Another consideration is that you can probably reduce correlation to other altruists' investments (I wrote about this a bit here, and I'm currently writing something more detailed). Uncorrelated investments have much higher marginal utility of returns, at least until they become popular enough that they represent a significant percentage of the altruistic portfolio. And leveraging uncorrelated investments looks particularly promising. So you could get more than a 1% excess certainty equivalent return that way.

Edit: Published Uncorrelated Investments for Altruists

Jonas Vollmer @ 2020-10-27T18:22 (+2)

Thanks, I look forward to your analysis of uncorrelated investments! In particular, I'll be keen to see to what degree they rely on the same assumptions as value/momentum strategies, or if there are opportunities that are independent of that.

jackmalde @ 2020-10-23T18:31 (+2)

Thanks for this very useful content!

A DAF can only give to legally registered charities. If you believe that, say, funding an early-stage startup is the best use of your money, then you can't do that if you put your money in a DAF

This seems a pretty big drawback to me. Obviously there are loads of great giving opportunities from an EA perspective that aren't registered charities. At the same time I do like the fact that DAFs can counter against value drift.

I wonder if there's scope for circumventing this issue by setting up a registered charity that can take donations from a DAF and then forward on to wherever the donor desires. Even an existing charity like CEA could act as a middle man like this. Is this a completely silly idea or a promising one?

MichaelDickens @ 2020-10-23T20:57 (+2)

I wonder if there's scope for circumventing this issue by setting up a registered charity that can take donations from a DAF and then forward on to wherever the donor desires. Even an existing charity like CEA could act as a middle man like this. Is this a completely silly idea or a promising one?

I'm not sure about this, but I don't think charities are allowed to give money to for-profits.

jackmalde @ 2020-10-24T06:24 (+1)

Thanks that makes sense.

I have just remembered about CEA’s EA Funds though which I understand essentially involves people giving money to CEA and CEA then forwarding the money onto various places, which aren’t necessarily other charities. Therefore it seems that my idea is possible? I don’t know if EA Funds actually gives to for-profits.

Do you know if it possible to give to an EA Fund from a DAF?

Larks @ 2021-05-28T05:46 (+10)

Do you know if it possible to give to an EA Fund from a DAF?

I was able to do this by disbursing to CEA and writing 'LTFF' in the comments.

MichaelDickens @ 2020-10-24T19:22 (+2)

Do you know if it possible to give to an EA Fund from a DAF?

That should definitely be possible.