Shots In The Dark
Friday, August 24, 2024
  Yale, Harvard Returns vs. Buy-and-Hold
A financial blog says the endowments could do just as well using a buy-and-hold strategy rather than the management systems the two universities have in place....
 
Comments:
Hi Richard,

I'm really out of my depth here, so I hope that you or one of your more financially savvy readers will correct me if I've got this wrong. But it seems to me that you've misread the financial blogger's post. He does not claim that the buy-and-hold strategy performs almost as well as the University endowments. Indeed, according to his chart the Harvard endowment has returned an average of 18.5% over the last five years while the B&H; strategy would have returned only 12.3%. (Harvard outperforms B&H; by over 25% since 1985, the beginning of the chart.) What is true is that the return streams are correlated at about 80% - which I take it means that when Harvard's return goes up the B&H; strategy is quite likely to go up as well. (Have I got this right?) This could happen, indeed *has* happened, even though the endowment far outperforms the buy-and-hold strategy. The strategy that he says actually performs (almost) as well as the two University endowments involves leveraging the portfolio by 50% and then investing according to a "market timing strategy" that he lays out in a linked paper. Perhaps that timing strategy is a lot less complicated than whatever strategy the endowment managers use - I have no idea. But glancing at the paper I can say that it seems to be a lot more complicated than simply buying and holding.

Sean
 
You're right; forgive my vagueness. That's why I said a "buy and hold strategy," but I can see that that phrase is slightly misleading....
 
Professor Kelly,

You have it exactly right in everything you said i.e. leverage, B&H;, timing. Each and every statement was correct.

Are you sure you're Professor Kelly of the philosophy department at Harvard and not the Sean Kelly at a particular mid- West hedge fund that I know?

I would add two other things to what you said. First, there is data mining going on here. In portfolio situations, in layman's terms, data mining means that one can almost always find a solution to most portfolio problems if it is done in hindsight. There are so many portfolios that can be constructed, that you can pretty much achieve anything that is reasonable. The problem is that what has worked in the past, doesn't always work in the future and extensive out of sample testing has to go on before one can begin to have any confidence in what one is doing..

Faber's thesis essential says that if we had invested in these asset classes, and if we had used a particular timing system, then we would have been able to replicate (not really, but I'll get to that in a second) the Harvard/Yale results during the last umpteen years.

Notice the use of the phrase (twice), "if we had." Clearly, we didn't, but on a "look back" it would have worked. Looking back, a lot of things would have worked. If I had bought stock in The Washington Post in 1975, I would have done very well. Had I invested in Microsoft at its IPO, I would have done very well. One could continue to go on and on. We can construct many portfolios that would have done as well. I have an even easier formula to achieve what Harvard has done. The benchmark at HMC until the early 90s was a 70-30 split between the S&P; and long term bonds. Leverage that up and it would have come close to the Harvard results. But then again, that is partially in hindsight.

The second thing is very simple. We could have done it if we had used leverage. I could have done a lot better with portfolios when I managed money professionally (e.g. the portion of Princeton’s endowment) if I had used leverage of 50%; I used none. However, in Faber’s case as one can see by the results, had leverage (and 50% at that!) not been used, then the results would have been far worse than Harvard’s. Put another way, if Harvard had used 50% leverage, it’s results would have been far better than the results Faber shows.

Faber compared apples and oranges. We see papers such as his all the time. They have no meaning in the real world of portfolio management.

Richard,
Sorry to have to tell you this but your reply makes no sense. What was it you were trying to say?
 
Sam,

Are you kidding? I know better than to tangle with you on this stuff.
 
Actually, what I think the blog writer is saying is that you as an individual investor could match Harvard's returns by using his proposed TAA strategy (tactical asset allocation).

For what its worth, TAA strategies have fallen out of favor and are pretty dimly viewed by most financial professionals. TAA still comes up against that nasty "Arithmetic of Active Management" rule as laid out by Bill Sharpe even if it is clad in an appealing clothes (i.e. the average investor will underperform the market after costs).

And Sam Spektor is absolutely correct that this is a classic example of data mining and rear view mirror analysis. I'm sure you could find a number of asset classes (commodities have done very well, for example) that would have outperformed the Harvard endowment, but that gives you no insight into anything.
 
Dear Sam Spektor,

Thanks for the helpful account. I'm glad you highlighted the leveraging and the retrospective analysis, since those are two things I wondered about. In particular, I meant to ask whether the endowment managers are allowed to leverage their portfolios. I gather that this is one of the ways hedge funds are able to do so well; if the endowment managers aren't allowed to leverage, as you say, then their recent performance must be pretty remarkable. Also, as you say, in retrospect it's pretty easy to find investment strategies that would have done pretty well. I suppose that if I'd put all my money in Microsoft back in the mid-80s then I'd have a pretty penny now. The trick is in knowing - back in the mid-80s - that that would turn out to be such a good strategy.

Sean

P.S. - I'm pretty sure that if I were at a hedge fund, mid-western or otherwise, I'd be making much more money than I am.
 
Professor Kelly,

Once again, very good questions. I'm impressed. It is not often that one gets a response like this by someone not in the field.

I’m not exactly sure who you were referring to when you said endowment managers. Were you referring the overall Harvard portfolio which is the investment and operational responsibility of Mohamed El-Erian who is the head of HMC, or some of the investment managers and hedge fund managers that Harvard uses?

I’ll give you the answers (or I’ll try to) for either situation. Either group can use leverage (I hadn’t realized I had said that endowment managers aren’t allowed to use leverage; I was referring only to the fact that I hadn’t used leverage in managing a portion of several endowments). Actually, in the case of Harvard, it typically has used a small amount (5% or so) of on balance sheet leverage (and, periodically, a very large amount of off balance sheet leverage for one internally managed strategy; that strategy is no longer managed internally). I might add that all of this information is available on line in Harvard’s financial report (don’t want to be accused, as I have been in the past, of using inside information). So… Harvard’s endowment as a whole typically uses a little leverage to enhance (hopefully) its investment returns.

In the case of Harvard (and Yale and other endowments), the outside money managers that they hire to manage funds (in Yale’s situation it is all managed externally; in Harvard’s partially managed internally, partially externally) can use and in most cases do use leverage (although my firm was the exception to the rule), sometimes exceedingly high amounts. Many times this works to the advantage of, let us say, the hedge fund and thus the endowments and others for which it manages money. Many times it doesn’t work as we’ve witnessed very recently (e.g. Sowood and others) and over the past 10 years (think blowup of LTCM in 1998). Leverage works both ways, up and down. Sometimes you have a good strategy that will work in the long run, but because of the high amounts of leverage, will turn out to be a losing strategy because it can’t weather the short run. For example, if a hedge fund is leveraged at say 7 to 1, if on a short term basis that strategy does not work and causes a decline of 15%, the entire equity capital is wiped out. A decline of 7 ½% results in a 50% decline. That is what apparently happened to some funds in the last month or two… the strategy was good long term but there were short term dislocations (unforeseen market conditions is the excuse generally stated; unforeseen, but happening with quite frequent regularity) which caused the strategy not to work, the negative leverage compounding the misery and bingo, the funds had huge losses. All of these funds forgot Maynard Keynes dictum… the market can remain irrational longer than you can remain solvent.

Hope this helps a bit, but if I’ve missed anything, haven’t made myself clear or can try to provide more information, please let me know.
 
Sam Spektor,

That was very helpful - thanks. I see now that you were clear in your original post that *you* hadn't leveraged your positions in the Princeton portfolio; I mistakenly took this to mean that University endowment managers in general don't or aren't allowed to do so. (I suppose I meant the internal fund managers at HMS, by the way, though in retrospect I realize I wasn't clear enough about the distinction.) In any case, thanks for clearing this up.

Sean
 
Post a Comment



<< Home
Politics, Media, Academia, Pop Culture, and More

Name: Richard Bradley
Location: New York, New York
ARCHIVES
2/1/05 - 3/1/05 / 3/1/05 - 4/1/05 / 4/1/05 - 5/1/05 / 5/1/05 - 6/1/05 / 6/1/05 - 7/1/05 / 7/1/05 - 8/1/05 / 8/1/05 - 9/1/05 / 9/1/05 - 10/1/05 / 10/1/05 - 11/1/05 / 11/1/05 - 12/1/05 / 12/1/05 - 1/1/06 / 1/1/06 - 2/1/06 / 2/1/06 - 3/1/06 / 3/1/06 - 4/1/06 / 4/1/06 - 5/1/06 / 5/1/06 - 6/1/06 / 6/1/06 - 7/1/06 / 7/1/06 - 8/1/06 / 8/1/06 - 9/1/06 / 9/1/06 - 10/1/06 / 10/1/06 - 11/1/06 / 11/1/06 - 12/1/06 / 12/1/06 - 1/1/07 / 1/1/07 - 2/1/07 / 2/1/07 - 3/1/07 / 3/1/07 - 4/1/07 / 4/1/07 - 5/1/07 / 5/1/07 - 6/1/07 / 6/1/07 - 7/1/07 / 7/1/07 - 8/1/07 / 8/1/07 - 9/1/07 /


Powered by Blogger