One thing which is clear, however, the institutions of that strength could, if they chose, apply to the Principal Supervisory Agent for the necessary approval to investment in direct investments if they chose; and if they are, indeed, strong, well-managed institutions, the likelihood of them being able to convince the Principal Supervisory Agent that they should be permitted to do so would be presumably high.

MR. SAHADI: Well, Harry, just by the fact that they have 6 percent tangible means that they can exceed 10 percent now without having to go to the PSA in twice their net worth.

MR. WHITE: Bob, I was going to bring it up at some point. You brought up this issue of portfolio, and it is one of the puzzles to me. In one of the studies that have been done, there is the apparent finding that the effect of direct investments for institutions--well, now, let me start back on this--it's to your question about risk.

If one is looking at an individual asset, then variance is a measure of risk, but for an institution, you care about the portfolio rather than individual assets, and so one has to ask about how an asset fits into the portfolio.

As a friend of mine has written, by itself, selling umbrellas looks pretty risky and the pattern of sales would vary from day to day. By itself, selling sun tan lotion would look pretty risky and the pattern of sales would vary from day to day.

Put them together, however, and you may have balanced out your sales portfolio with the variance in the one offsetting the variance in the other; and the claim in one of the studies is for institutions at direct investments of above 5 percent and above 10 percent. The direct investments have the effect that I just described, that they tend to balance out the portfolio to reduce the overall variance of the portfolio rather than exacerbate it.

And I'm interested in what reaction you have on this.

MR. SAHADI: I would certainly agree with that theory, that I think does not argue against this rule that we are saying by the fact that we have a 10 percent threshold and even approval above that, that our supervisory agents aren't looking for a salutory portfolio effect in relation to this.

If you want to talk to the direct study that talks about direct investment in these portfolios and our analysis of that, I'll turn it over to Don Bisenius.

MR. BISENIUS: It's always enjoyable to be on the other side of the study rather than defending one, to be able to criticize one. The study we're referring to, again, was one done by Professor Benston who attempted to analyze the diversification. At least one aspect of the study was to look at the diversification opportunities afforded direct investment. His results indicate that, on average, institutions have received some diversification benefits from their involvement in direct investment. I find that result interesting and, I guess, consistent with what theory would suggest, but I think there has to be some caution in interpreting those results, and, I, at least noted three that somewhat bothered me.

In coming up with his returns from which he derived variance measures, the returns were averaged over three years. As you're aware, anytime one averages returns over years, you reduce the overall variance and therefore whether or not the variance in correlation measures are accurate or reflective of the underlying asset is, at least, suspect.

Secondly, the returns that are investigated do not take into account or at least potentially do not take into account, capital gains and loses. That's a point that Mr. Black has referred to earlier. They look at accounting income. To the extent there are capital gains and losses on the project, those should somehow be distributed over the various years of the project, which will influence both the returns and potentially the variance.

The third issue, and again, this gets, I guess, somewhat technical, and I think an important aspect of it, has to do with the variance measures that were used in this study, and that is, what was used was a crosssectional variance. You took all the direct investments. You looked at their returns and you saw how much variability there were in returns across direct investments. That doesn't seem the appropriate variance measure to

use.

If one wants to look at the variance of a direct investment, one would prefer to have a time series on that investment, to be able to know, how does this one project vary over time. The analogy, the example, I think that comes to mind is, if I was going to invest in GM stock, I wouldn't necessarily look at the return on GM stock, Ford stock and Chrysler stock, calculate the variance and say that's the riskiness of GM stock; rather I would look at GM stock over time and see how much fluctuations there were in holding that stock in my portfolio.

So, I'm not sure that the variance measures that were used in these studies accurately reflect the riskiness of the projects, and therefore, inappropriate in determining whether or not it provides the diversification benefits.

If those criticisms can be dealt with, then, in fact, his study is valid. It provides diversification opportunities.

MR. WHITE: Thank you, Don.

I think next time we have a Board meeting we ought to talk about heteroskidasticity, simultaneous equations, bias and sufficient estimators.

(Laughter.)

MR. GRAY: Maybe you can explain what all that means.

MR. WHITE: Frank, we're talking about the problems of high flyers, et cetera. As you know, a number of the commenters said, yes, that was a problem in the past, but the new capital regulations are going to bring that under control because the high fliers are going to have to find the money somewhere. And, again, once they start putting up their own money, do we have to work? Do you have any sense of what the new capital regulations are?

MR. PASSARELLI: One thing about a high flier, he gets his money back one way or the other. I mean, in other words, these institutions get organized and they'll actually come in with systems. So, that they were able to get their investment out of this institution before it goes broke. I mean, that's usually what our experience has been. What they 11 do, they'll actually provide some means of getting some kind of fee income. So, the Net Worth Regulation will not actually be the controlling factor. mean, that of itself is only one aspect of a control.

I

If people want to provide additional net worth, that, alone, does not provide the necessary protection, in my view, because what happens is these people, they will actually generate the necessary income with some of these high-risk type of activity, and as a result, they'll have the income and also provide a way for them to get their investment out of these institutions before they become insolvent.

MR. WHITE: Yeah, but doesn't that, in a sense, go too far? I mean, that almost says, you got these high fliers out there, they're going to get their money out no matter, and if it isn't direct investment, they'll go out and hedge in improper ways rather than proper ways and, or, you know, leverage themselves in some way that even I couldn't even begin to think about.

MR. PASSARELLI: I don't disagree with it.

MR. QUILLIAN: There's another kind of set of considerations there, if I may, Board Member White. Our allusions to the new regulatory capital regulation, of course, sometimes overlooked the fact that is a new regulation, and the increased capital strength which it looks to will be built up over time, as much as six to twelve years into the future.

And, so, as the institutions come into compliance with the regulation as it takes hold, and gains strength, well, that, of course, will be more relevant. That brings to mind, of course, a number of considerations, and that is, that if FSLIC itself were stronger; if the industry had stronger real capital of its own; if there were fewer institutions in rather desperate conditions than there are now; then one might take a bit more laid back attitude towards the dangers which we believe inhere in excessive direct investment.

But of course, those conditions don't obtain. We do have several hundred institutions that are in rather desperate conditions, we have a severely undercapitalized FSLIC. We have an undercapitalized industry which we regulate, and this regulation addresses that current condition.

MR. BLACK: Everyone on the staff agrees that increased real capital is beneficial and salutory in lots of different ways, and this is one of the fields salutory. Aside from the phasing, of course, the current regulation doesn't require tangible net worth, and it has limitations in terms of deterring, plunging, inherent, and not going to a tangible net worth

measure.

In terms of the question to Mr. Passarelli, there is that danger that they'll do something else, but most of even our worst shopped attempt to follow the regulations, they attempt to do so because it slows down enforcement efforts against them.

And what we see time after time is folks basically gaining their net worth to be able to do what they wanted. Empire reported record profitability until the very verge of collapse, and there are lots of ways to do that, including under our new capital requirements and to burst off into very large measures of direct investment or other risky investments and, as I read from this January '85 one, it isn't just direct investment. No one is claiming it is just direct investment. The same folks who tend to put very heavy concentrations in their portfolio of direct investment, also choose very risky other areas in their portfolio, in terms of commercial loans, ADC loans, of very volatile deposit base. They are plungers.

MR. WHITE: You just, again, I think, reenforced my concern with, for example, the California study that just by looking at the data in front of me, I don't know what really was causing failure.

MR. BLACK: And we understand your point. We understand about the limits of analysis, and particularly with low sample sizes, and I get back to again, since--yes, insurance companies try to do it right, but insurance companies try to stay in business first, and when they see their loss experience, they don't wait until they can have the most perfect study in the world. If they see they're taking losses up the kazoo in those areas, they cut it off.

MR. GRAY: But let me go a step further here. The House Government Operations Committee report that came out, I believe, on November 5th of 1985, went into this whole subject in great detail over, I think, a matter of something like five or six months. And I think there was general recognition that there wasn't nearly as much data as they would like. Let me read to you what they concluded. They said that "As long as the FSLIC fund remains impaired" -- and no one can doubt that it remains impaired-"as long as the FSLIC fund remains impaired, the Federal Home Loan Bank Board's direct investment rule is an appropriate and necessary restriction."

And there was another part of the report which called the direct investment rule "a prudent precaution while the FSLIC is in a weakened condition." That's what we're really talking about here.

What kind of reasonable precautions are we taking when the FSLIC fund is in a weakened, and remains in a weakened condition, more weak today than ever before in my memory?

MR. BLACK: I think it gets back to one of the questions Board Member White asked. He says he wants to make a rational decision, which everyone agrees with, which involves, let's look at the net effect of direct investment. I see here evidence in what it causes in terms of failures, but let me look at the up side, too.

Let's look at the up side because I agree, and, particularly if you are distinct from an insurance company analogy, but even if you just talk about it from a government regulator's perspective, look at the up side.

Who was above 10 percent? Remember if you're below 10 percent, you can do it anyway, and we're not stopping you from doing it. So, our reg doesn't hurt you, doesn't stop you from that up side.

Who was above 10 percent? It's not a sample. It's the universe, these 37 institutions that were above 10 percent. What up side harm could we possibly have done if three years ago we had prohibited those institutions from going above 10 percent in direct investments? That's the up side that you might be losing.

I submit, if you look at it in those terms, it gets to be a relatively easy decision. For any insurer, it's a super-easy decision, but I think even academically, if you look at it on that up side versus the down side from the FSLIC cost numbers, it doesn't get to be that hard a decision.

MR. QUILLIAN: There is, of course, Bill, the additional consideration that we're not talking about a regulation which is a ceiling or a firm limitation or a prohibition, but rather a regulation which establishes a threshold for supervisory review.