Every investor (and every fund manager) knows that fees matter. Funds with higher fees tend to underperform funds with lower fees (after fees that is). But fees as a share of assets under management are not the only fees levied on investors. Especially hedge funds tend to levy performance fees as well and while they are easy to understand in theory, the details can have a significant influence on investment returns.
A team from University College London ran millions of simulations for hypothetical hedge funds that charge a flat fee plus a performance fee to see how $1 invested in these hedge funds would be worth after ten years. The beauty of these simulations is that one can change the volatility of the hedge fund, the size of the performance fee, the existence of a hurdle rate or a high-water mark, and the frequency with which fees are charged by the fund. Three results are well-known or at least should be well-known to investors:
The more volatile the return of the fund, the higher the effective fee for the investor. This happens because a more volatile fund is more likely to have higher returns in one year, triggering higher performance fees, followed by lower fees the next. But negative performance means no performance fees no matter whether the fund loses 1% or 10%.
High-water marks reduce the above effect because it takes longer to recover from bigger losses and collect additional performance fees again.
Similarly, a hurdle rate can reduce this effect if it is set not as a fixed hurdle (e.g. 5%) but more in relation to the performance of a reference index like the S&P 500.
The one result that is commonly not appreciated enough by investors is the impact of the frequency with which fees are charged. Some funds charge their fixed and performance fees every quarter, some do it once a year. But this makes a difference, particularly in adverse environments when the fund has negative returns.
The chart below shows the worst-case outcome of the simulated development of $1 invested in funds with variable volatility but a 20% performance fee and a high-water mark.
Worst case outcome for $1 invested with 20% performance fee and high-water mark
Source: Galas et al. (2024)
As we can see, if fees are levied every quarter instead of every year, the assets can shrink much more because fees eat up more of the assets over time. The same is true if the performance fees are subject to a hurdle rate of 5% rather than a high-water mark.
Worst case outcome for $1 invested with 20% performance fee and hurdle rate
Source: Galas et al. (2024)
Note that as a general rule, investors who are charged annually end up with more money than investors who are charged quarterly. The effect gets larger the more volatile a fund’s returns are. For funds with a volatility of 10% to 20%, the difference can easily reach 10% to 20% of the money lost due to nominally identical fees but charged at a higher frequency.
What happens here is essentially the eighth wonder of the world in reverse. If fees are charged more often, capital cannot compound at the same rate because any recovery in funds is interrupted by the fees being charged. Over time this is the same as consuming dividend income in a stock instead of reinvesting the dividend into the stock. And this small difference grows larger and larger the longer that happens.
trading as much as he did. Sigh. [Note that my first reply was cut off, and I couldn't edit it for some reason ... sorry.]
I would believe a similar effect would be observed in systems that tax reinvested dividends as "qualified" (meeting a certain holding period threshold, so taxed at a long-term capital gains rate rather than as short-term/ordinary income), or not at all (as with IRA/401(k)/ISA/superannuation retirement accounts).
There's another great industry unspoken: One would suppose that more volatile funds would have a higher propensity to trade (either reacting to or perhaps even exacerbating the volatility), which would drag on returns even further. I once met a fund manager who said he was unconcerned about institutional transaction costs because they were buried in the broker bid/ask spread and thus "out of his control"; I replied that they certainly were, as he could refrain from trading so much. Sigh.