Why would anyone invest in these funds?

[Maverick1]

Twenty years of data is barely enough to show that there is statistical evidence that Wintons returns are positive. However they could plausibly be just above zero or over one.

Ten years of data tells us basically nothing. So actually it's a moot point whether the estimated SR of Cantab was 0.28, 0.5 (in which case the range would be -0.17 to 1.17) or something else

So, if twenty years of data on returns is not enough, and 10 years of data on returns basically tells us nothing, what do 10 and 20 years of data tell us about correlations?

** Not that it matters but I hand calculated the SR for Cantab in my head and obviously got it wrong, for the time period shown it is 0.28. However there is about a year of missing data so I don't actually know what the latest estimate for Cantabs SR is.

Thanks.

 

[Maverick1]

The linked page from your post shows:
Annualized Return 7.50% Annualized Std. Deviation 17.92%
Which, using rf== 0%, would be 0.42 Sharpe, close to GAT's estimate.
(The linked page uses rf== 2½%, which has a big downward impact on small Sharpes)

Reminds me of a t-shirt once seen on a stats professor: "When all fails, manipulate the data"

Lol.

Jes messin with ya

 

[truetype]

It's a matter of taste. I always use rf== 0% because I don't want to spend brain cells on backing out this-or-that rf rate. Similarly, I use pre-slippage numbers.

 

[globalarbtrader]

So, if twenty years of data on returns is not enough, and 10 years of data on returns basically tells us nothing, what do 10 and 20 years of data tell us about correlations?




Thanks.

Excellent question. There is no easy closed form, but when I looked at the correlation of US stocks and bonds (which is also slightly negative) after five years the 90% range was about -0.3 to -0.2; and the range doesn't get much tighter after that.

So in simple terms we need relatively little data to form an opinion on correlations, and we can be much more confident about our estimates of correlation than we can about Sharpe Ratios. Incidentally the same is true of volatility.

If the estimated correlation of Cantab and SP500 was -0.1 then after ten years the range would be around -0.15 to -0.05.

GAT

 

[Maverick1]

Excellent question. There is no easy closed form, but when I looked at the correlation of US stocks and bonds (which is also slightly negative) after five years the 90% range was about -0.3 to -0.2; and the range doesn't get much tighter after that.

So in simple terms we need relatively little data to form an opinion on correlations

Huh?

 

[globalarbtrader]

Huh?

Sorry, which bit isn't clear?

GAT

 

[Maverick1]

It's a matter of taste. I always use rf== 0% because I don't want to spend brain cells on backing out this-or-that rf rate. Similarly, I use pre-slippage numbers.

Wouldn't a conservative approach (rf = 10y) be more prudent, over and against convenience, especially considering that rates have been heading higher?

 

[Maverick1]

Sorry, which bit isn't clear?

GAT

All of the post.

 

[globalarbtrader]

All of the post.

This isn't clear?

In simple terms we need relatively little data to form an opinion on correlations, and we can be much more confident about our estimates of correlation than we can about Sharpe Ratios.

If the estimated correlation of Cantab and SP500 was -0.1 then after ten years the range would be around -0.15 to -0.05.

I'm not trying to prove anything, I'm just answering your question (what the statistics are for correlation estimates).

GAT

 

[Maverick1]

This isn't clear?

In simple terms we need relatively little data to form an opinion on correlations, and we can be much more confident about our estimates of correlation than we can about Sharpe Ratios.

If the estimated correlation of Cantab and SP500 was -0.1 then after ten years the range would be around -0.15 to -0.05.

I'm not trying to prove anything, I'm just answering your question (what the statistics are for correlation estimates).

GAT

You are assuming your conclusion, once again begging the question on this claim above re "we need relatively little data". You also contradict your own statements below. If 10-20 years is not enough for statistical knowledge on returns, it is also not enough for statistical knowledge of correlations. Ask LTCM.

Twenty years of data is barely enough to show that there is statistical evidence that Wintons returns are positive. However they could plausibly be just above zero or over one.

Ten years of data tells us basically nothing. So actually it's a moot point whether the estimated SR of Cantab was 0.28, 0.5 (in which case the range would be -0.17 to 1.17) or something else

And btw, do yourself a favor, and be careful about throwing words like "daft" out next time. You never know what's going to come back and hit you.