24 Jul 2016

The Problem of Compounding Fees

Compounding is magical. When you invest in a mutual fund, you get a return, but that return gets reinvested and earns additional return.

Imagine someone who starts earning at the age of 23, and starts investing a rupee a month. He chooses a mutual fund that gives 12% return [1]. He does this till the age of 60, at which point his corpus will be ₹7000 [2].

Suppose the interest was simple rather than compound. In other words, suppose each month’s investment earned interest for the next one year but not after that. Then the final value will be less than ₹500.

What a difference between ₹500 and ₹7000! Almost a 14x difference. This is all due to the magic of compounding.

But the other side of the coin is compound fees. It’s not just your returns that compound but also the fees you pay. In the above example, suppose the investor picked a mutual fund that has a low annual expense ratio of 1%. Which means the annual return is now 11%. This reduces the final portfolio value from ₹7000 to ₹5400.

That’s a loss of 23% in the final portfolio value [3]. But the fee was only 1%! How did 1% become 23%? It compounded, like your return compounded.

See how unintuitive compounding is. As Einstein put it, compound interest is the eighth wonder of the world. He who understands it, earns it. He who doesn’t, pays it.

What can we do about this?

One obvious answer is low expense ratios like 0.3% [4], as in the US.

Another is to have a non-compounding fee. Let’s say a fund levies a non-compounding fee of 1%. When you invest, 1% is deducted. And when the value of your investment later increases in value because of NAV increase, only 1% of the increase will be deducted as fee [5].

Suppose you make an initial investment of 1 lac. 1K is deducted as fee, leaving you with 99K. After a year, suppose your investment increases in value to 1.1 lac. Then, your annual expense will be 1% of the increase of 11K, because you already paid the fee on the principal of 99K. So you need to pay only ₹110. As opposed to the status quo, where you pay ₹1100. That’s a 10x difference in fee!

One step better than a non-compounding fee is to completely get rid of the annual expense, and have only entry and exit loads. The entry load applies to everyone, and the exit load only to people who redeemed their investment before a decade. Unlike exit loads today, which are required to be distributed to the remaining investors, the fund house should be allowed to keep them. I don’t mind them being high, like 10%.

This will also reward long-term investors, while incenting other investors to invest for the long-term.

A fund with a 10% entry load and no other fee is similar to one with a 10% non-compounding fee, in that in both cases you’re charged 10% of your initial investment when you make it, but in the non-compounding fee case, you’re also charged 10% of the increase in value of your portfolio.

In conclusion, we should have funds with lower fees compounded over the decades you invest. This could be achieved by a lower annual expense ratio, a non-compounding fee, or replacing the annual fee with entry and exit loads.

[1] Funds don’t give the same return every year, which makes this calculation illustrative, not exact.

[2] All numbers are rounded off for readability.

[3] (7000 - 5400) / 7000. You can also calculate the loss as a percentage of 5400, which gives an even higher result, but I choose a more conservative measurement. Even with a conservative measurement, you’re losing a lot.

[4] If the concern is that insufficient money will be available for research or trading, one answer is to see if that research and trading really generates more return than it costs. We won’t know unless we can compare for ourselves, rather than listening to the industry’s self-serving claims. Let there be a dozen funds with an annual expense ratio of 0.3%, so that we can see if lower fees work in the investor’s favor.

Another answer is to have the industry consolidate into fewer funds with greater assets under management each, so that a smaller percentage of a bigger number is still more or less the same amount. That is, expense ratios can decrease from 1.5% to 0.3%, which is a 5x decrease, if AUM increases 5x to compensate.

[5] What if the NAV increases, then declines and then increases? Will you pay a fee twice on the increase? That is, say the value goes up from 1 to 1.1 lac, at which point you pay a fee based on the increase of 10K. Then it declines back to 1 lac. And then increases again to 1.1 lac. Will you pay a fee again on the increase of 10K? Probably not. Maybe the fee should be calculated based on the high-water mark of your balance. Which is different for different investors.

No comments:

Post a Comment