Entrevista David Swensen Parte 01 - Alocc Gestão Patrimonial

So when I began at Yale, actually 29 years ago on April 1st, and I found out that there's April Fools' Day in Brazil, so we're trying to figure out exactly where the joke is. I looked around at how other endowments were managed, and there were, I think, two fundamental problems. One was lack of diversification. Half of the assets were in U.S. stocks, and 40% of the assets were in U.S. bonds and cash, and 90% in U.S. marketable securities. And if there's anything that we learn from finance theory, and if there's anything we learn from common sense, diversification is important. And so one of the most critical components of what people call the Yale model is to have a well-diversified portfolio. And of course, the way that you

express that diversification is through your asset allocation analysis. And you have to identify a number of asset classes, five or six or seven or eight, something like that, and then spread the assets around those asset classes so that each of them has a meaningful impact on the portfolio, but no one has an overwhelming impact on the portfolio.

One more question regarding this strict topic. It's about the differences of a portfolio of a family, a family office portfolio, and an endowment portfolio. Do you see much differences? Sure, please consider size. Let's talk about a hundred million portfolio of a family and billions of an endowment.

So I think the most fundamental question is time horizon. And so if the family portfolio is substantial enough that it is to satisfy the needs of several generations, then I think that there's a remarkable coincidence between the investment challenge faced by an endowment and faced by a family. I think that the types of strategies that family offices and endowments pursue are often very aligned. You're right to mention size as being an issue, but of course there are families that have assets that are much greater than Yale's $23 billion portfolio. There's no difference at all. And then there are smaller endowments as well that are $50 or $100 million. And so there is a difference in terms of how you manage smaller portfolios as opposed to larger portfolios.

But generally speaking, I think, adjusted for size, the challenges faced by family offices are very similar to the challenges faced by endowments.

For example, it could well be that bonds are priced to give you inflation plus five today, but what are they gonna give you next year? You don't know. I don't know either. Nobody knows, right? And if you say, well, I'm perfectly happy just to earn inflation plus five, and all of a sudden bonds aren't priced to give you those kinds of returns, then you're stuck. So that doesn't mean that it's not a relatively attractive asset, and it doesn't mean that you might not want to have a substantial allocation to inflation-linked bonds, but my mentor at Yale was Jim Tobin, who won the Nobel Prize in Economics. And Jim was asked to explain to a New York Times reporter what it was that he won the Nobel Prize for.

And he said, well, I guess you could say, don't put all of your eggs in one basket. It's a universal figure. It's a universal figure. So we all could have won the Nobel Prize, except I think Jim had some fancy math to go along with it. But even if something looks very attractive, don't put all of your eggs in one basket. So why don't I start out by saying how it is that I don't define risk, at least not most fundamentally? And that's in terms of, you know, statistical variation in returns through standard deviations or variances. I think Warren Buffett once remarked that if a stock price went down dramatically and then you recalculated the statistical risk, you'd say, oh, the risk has gone up substantially because of this dramatic one-day fall in the stock price.

But he would say that because it's cheaper, all other things being equal, the risk is lower. And I think that there's a lot of common sense in that attitude. I like to think about risk more as the chance that you will have a fundamental impairment in value of your assets. And that's one of the reasons why I'm very wary of leverage in the portfolio. That doesn't mean that leverage that you find on corporate balance sheets or sensible amounts of leverage used in buyout transactions can't be a sensible part of the overall portfolio. But when people take portfolios of assets and add unreasonable amounts of leverage, then I think you're exposing the assets to levels of risk that fundamentally don't make sense.

And so in those instances, you might not see it in the standard deviation of returns, but you certainly are putting the fundamental value of the assets that you're responsible for at risk, and I don't like that.

And what about, I mean, the illiquid assets, they're risky, they're more risk?

Oh, on the illiquid side?

Yes.

So... there are a couple answers to that. I think if illiquidity is limited to 20 or 25 or 30% of the portfolio, that by and large, that doesn't pose any particular challenges. But if you have a portfolio like Yale's, where as much as 50 or 60% of the assets are illiquid, then you have to make sure that you've got plenty of sources of liquidity, not only to support the investment management requirements that come along with illiquidity. For example, if you're in private partnerships, the partner can call funds and you have to send them funds when the calls are made. But in the case of Yale, we're now sending more than a billion dollars a year to support the operations of the university to pay for faculty salaries, support research, pay for student scholarships.

And that's not—it's not flexible; it has to happen. And so you need to make sure that you've got enough liquidity to support both the investment operations and your responsibilities to the university. In the standard mean-variance framework that people use to analyze portfolios, liquidity doesn't show up. It's a judgment factor that you need to apply. Not that you just follow the mean-variance optimization precepts anyway. I think they're useful tools and interesting tools. You always have to have a combination of the best analysis that you can put together and the best-informed market judgment that you can bring to bear. And the liquidity considerations are part of the informed judgment. It doesn't show up on the analytical side using the standard tools that we've got available to us.

So those are tough questions. And I think some of the characteristics that you described for real estate are absolutely true. But if you're doing it with a relatively small portfolio, there are also some very substantial risks that go along with direct investments in real estate. I mean, first of all, just as in the discussion with the inflation-protected securities, you should limit your exposure to any particular asset class. So even though real estate has all these potentially very nice characteristics, it should still be one of a number of asset classes in the portfolio and you should limit your exposure because you just don't know what the future brings. I think one of the problems that can come from the circumstances you describe is that people tend to talk about their successes

and they don't talk about their failures.

Exactly.

And I would suspect that most people that are managing real estate directly on a small scale aren't doing the same kind of portfolio analytics that you do when you look at the family office structures that you've put in place. But one of the risks that occurs to me is that even though in aggregate real estate may do well, the individual location that you're talking about with a specific building is incredibly important. And if you can't diversify over 10 or 15 or 20 different properties and you've got two or three and they turn out to be locations that don't end up being as attractive as they might have been in the first instance or as the entrepreneur might have thought that they were going to become, then you can end up with something that has far less value than you might have thought.

These real estate characteristics that you're talking about depend on the markets being in equilibrium. And if you end up with a real estate parcel in a location for which there's no demand, then you're stuck, right? And more generally, how do you think about the risk characteristics of real estate? Well, if you're talking about something like an office building that's well-located with long-term leases, then it probably looks like something between a stock and a bond because you've got equity risk, particularly with the residual. But if you've got leases to reasonably high-quality tenants, that looks like a bond. So the risk and return characteristics should probably be somewhere between bonds on one end of the spectrum and stocks on the other end.

Yeah, but you have against you depreciation, the movement of the cities. How do you compare the volatility of real estate under a REIT and a normal stock?

I think it's really hard to generalize about real estate. I was talking about this with a friend the other day. Twenty years ago, we thought the big regional malls in the United States were one of the highest-quality assets that you could possibly have. And they had quasi-monopoly positions in the region and high-quality tenants and increasing demand. And I think this is, order of magnitude, right. Over the last three years, foot traffic in malls in the United States has gone down by 30%. Right, so that's a fundamentally different set of risks and concerns as compared to central business district office space, right? Which has held up remarkably well in the major cities, which is, again, fundamentally different from suburban office,

which now seems to be at a disadvantage because people are moving back to the cities. So it's very difficult to generalize because there are all these cross-currents between the various sub-asset types within real estate. I think it should be part of the portfolio, but it should be a limited part of the portfolio because there are material risks, but, of course, there are in all aspects of investment.