5 ms·

Context is important, so the Nasdaq 100 (as represented by the QQQ ETF) returned ~49% in 2020, 39% in 2019, and 24% over the last 5 years. Beating the indexes b

by runako 6y ago

Context is important, so the Nasdaq 100 (as represented by the QQQ ETF) returned ~49% in 2020, 39% in 2019, and 24% over the last 5 years. Beating the indexes by 5-6 points consistently is very good, but it's important to keep in mind that most equities were doing really well over that period. (Edit: It's been reported elsewhere that these numbers are net of fees. However, it's entirely possible that for taxable accounts, tax considerations still narrow the performance gap. And of course, the QQQ offers instant liquidity and did not drop 53% this month.) Either way, the money currently in there is very likely to underperform the indexes over the next 5 years (because they need to return 100% to get back to the pre-GME position).

dcolkitt 6y ago
The correct benchmark for a hedge fund is T-bills, not the S&P or the Nasdaq. That's because hedge funds are an absolute return product that offers an income stream uncorrelated to the market. This may sound counterintuitive, but it's the basis of modern portfolio theory. The price that an investor should be willing to pay for an investment has to do with its beta to the broader market. Think of it this way, imagine you could access the S&P 500 in a parallel universe. It has the same return characteristics as the normal S&P 500, but in any given year moves independently. How much does this improve your portfolio? Intuitively you'd think it's not worth anything. It's only as good as your current investments, so what's the point. But in fact modern portfolio theory tells us that it's a huge improvement. Investing 50/50 in S&P and Bizarro-S&P, substantially improves the amount of return you can access for the same risk. That's because the two diversify each other, and the blend either reduces risk by 30% or lets you leverage up and increase expected returns by 30%. You're saying, "maybe Melvin isn't a good investment, because it seems about equal to the S&P". But the point is an investment that's about equal to the S&P, yet uncorrelated to the S&P is massively valuable. Even if it has big drawdowns, you typically don't care as long as those drawdowns tend to occur during times when the rest of your portfolio is doing fine. This is the same reason that a 60/40 stock-bond portfolio has massively outperformed 100% stocks historically. Even though bonds themselves return less than stocks.
- letsgoyeti 6y ago
  Thanks for the explanation - one question though, how do we know that Melvin (or any other security/fund) is not correlated to the S&P?
  - dcolkitt 6y ago
    To be fair, it probably is. Most equity hedge funds are probably closer to a 50% correlation, and should be benchmarked accordingly (i.e. half T-bills/half S&P). Investors would be able to see its historical monthly returns and regress them against the equity index.
- runako 6y ago
  This is good background and all makes sense, but in the context of Melvin we don't have enough data to know whether any of this is true in their case vs just being part of their pitch. (This is essentially true of all hedge funds.) What appears to be the case is they were edging the QQQ until they incinerated half their capital in a week on a bad bet and poor risk controls. So we now know they are uncorrelated with the S&P, but unfortunately we only know they are uncorrelated to the downside. (We don't know if their outperformance has been a result of real alpha or a strong market plus leverage.) We saw a lot of these funds in the strong bull market of the 90s and the takeaway I remember was that it's impossible to know ahead of time who's going to blow up and lose all/most of your money.
- koonsolo 6y ago
  > Investing 50/50 in S&P and Bizarro-S&P, substantially improves the amount of return you can access for the same risk. I can see how this would lower the volatility of your portfolio. But how do you improve the return in this scenario? > This is the same reason that a 60/40 stock-bond portfolio has massively outperformed 100% stocks historically How is this possible? If I invest 100% in stocks, my return after 10 years is higher than 60/40 stock-bond. So why do you say the latter outperforms it?
  - londons_explore 6y ago
    > I can see how this would lower the volatility of your portfolio. But how do you improve the return in this scenario? Leverage. You can translate reduced risk into greater returns by getting loans to invest more.
weeboid 6y ago
Doesn’t this 50% loss apply to the cum sum of gains though? That is, if I started with $1 and made $1,000, a 50% loss would still be $500
- runako 6y ago
  Yes you have more money than you started with, but you will still be down compared with an index that did not have a 50% drawdown at any point. - Unless you were late to the party. If you didn't get in the fund until (say) November, you have lost money.

by runako 6y ago

dcolkitt 6y ago

The correct benchmark for a hedge fund is T-bills, not the S&P or the Nasdaq. That's because hedge funds are an absolute return product that offers an income stream uncorrelated to the market. This may sound counterintuitive, but it's the basis of modern portfolio theory. The price that an investor should be willing to pay for an investment has to do with its beta to the broader market. Think of it this way, imagine you could access the S&P 500 in a parallel universe. It has the same return characteristics as the normal S&P 500, but in any given year moves independently. How much does this improve your portfolio? Intuitively you'd think it's not worth anything. It's only as good as your current investments, so what's the point. But in fact modern portfolio theory tells us that it's a huge improvement. Investing 50/50 in S&P and Bizarro-S&P, substantially improves the amount of return you can access for the same risk. That's because the two diversify each other, and the blend either reduces risk by 30% or lets you leverage up and increase expected returns by 30%. You're saying, "maybe Melvin isn't a good investment, because it seems about equal to the S&P". But the point is an investment that's about equal to the S&P, yet uncorrelated to the S&P is massively valuable. Even if it has big drawdowns, you typically don't care as long as those drawdowns tend to occur during times when the rest of your portfolio is doing fine. This is the same reason that a 60/40 stock-bond portfolio has massively outperformed 100% stocks historically. Even though bonds themselves return less than stocks.

letsgoyeti 6y ago
Thanks for the explanation - one question though, how do we know that Melvin (or any other security/fund) is not correlated to the S&P?
- dcolkitt 6y ago
  To be fair, it probably is. Most equity hedge funds are probably closer to a 50% correlation, and should be benchmarked accordingly (i.e. half T-bills/half S&P). Investors would be able to see its historical monthly returns and regress them against the equity index.
runako 6y ago
This is good background and all makes sense, but in the context of Melvin we don't have enough data to know whether any of this is true in their case vs just being part of their pitch. (This is essentially true of all hedge funds.) What appears to be the case is they were edging the QQQ until they incinerated half their capital in a week on a bad bet and poor risk controls. So we now know they are uncorrelated with the S&P, but unfortunately we only know they are uncorrelated to the downside. (We don't know if their outperformance has been a result of real alpha or a strong market plus leverage.) We saw a lot of these funds in the strong bull market of the 90s and the takeaway I remember was that it's impossible to know ahead of time who's going to blow up and lose all/most of your money.
koonsolo 6y ago
> Investing 50/50 in S&P and Bizarro-S&P, substantially improves the amount of return you can access for the same risk. I can see how this would lower the volatility of your portfolio. But how do you improve the return in this scenario? > This is the same reason that a 60/40 stock-bond portfolio has massively outperformed 100% stocks historically How is this possible? If I invest 100% in stocks, my return after 10 years is higher than 60/40 stock-bond. So why do you say the latter outperforms it?
- londons_explore 6y ago
  > I can see how this would lower the volatility of your portfolio. But how do you improve the return in this scenario? Leverage. You can translate reduced risk into greater returns by getting loans to invest more.

weeboid 6y ago

Doesn’t this 50% loss apply to the cum sum of gains though? That is, if I started with $1 and made $1,000, a 50% loss would still be $500

runako 6y ago
Yes you have more money than you started with, but you will still be down compared with an index that did not have a 50% drawdown at any point. - Unless you were late to the party. If you didn't get in the fund until (say) November, you have lost money.