Social Insurance

What follows is part of a discussion concerning notions of "social insurance," prompted by discussion of whether a "dividend tax" is a "good thing" or not . . .

==============================

. . .

Maybe I'm wrong, but I've been thinking we've been (or at least started . . .) talking about the dividend tax, which is not a tax paid by the corporation, but is rather a tax paid by the investor. My suggestion was not that the corporation was being "bailed out" but rather that investors were being (in some sense) "protected against risk." While corporate decision-making may well be affected by the existence (or level) of dividend taxation, that doesn't make it a tax paid by the corporation . . . But anyway.

Let me see if I can open out some of the general issues I was gesturing at with the term "social insurance," and why I raised in particular the examples of S&Ls and LTCM. I wasn't talking per se about a traditional "directly purchased" insurance policy . . .

Consider the S&L "bailout." As you pointed out, savings accounts in S&Ls at the time of the bailout process were insured up to $100,000 (in theory, via "premiums" paid to the FSLIC). But, there were some contributing factors to the "need" for a bailout. In the early 80s, the insurance limit was raised (from $40,000 to $100,000), and S&Ls were significantly "deregulated" with respect to lending practices and ownership. Thus, S&Ls became a wonderful "risk reduced investment opportunity," both for depositors and for other investors (e.g., owners of S&Ls). S&Ls "boomed," far outstripping the reserves FSLIC had to cover losses. Beyond that, the deregulation of the S&Ls opened the doors for S&Ls to engage in significanltly riskier investments of their resources. S&Ls lost tons of money, and were failing left and right. As it went, FSLIC itself went "bankrupt" (i.e., couldn't cover the losses). Hence, congress ended up commiting taxpayers to bailing out the failing S&Ls. There were two general mechanisms for the bailout. The first was (interestingly, given recent activities in California?) the issuance of long term bonds (30 year bonds -- we are still paying them off today). The second was the creation of a Resolution Trust Corporation, to manage the disposal of assets of the failed S&Ls. The RTC essentially held a "fire sale" (see below on LTCM . . .) of the assets, at significantly below market value (partially with the intention of "stabilizing" the real estate market . . .). As you noted, taxpayers were the "lender of last resort" tagged with the bill.

The net effect was that depositors/investors in S&Ls enjoyed protection against risk -- in effect, a form of "social insurance" . . . (Ummm, I followed the S&L process, with some dismay, relatively closely -- in the early 70s, during a five year hiatus from the halls of academe, I worked at an S&L in the accounting division, primarily as the "budget/forecasting analyst," so I had a personal interest. The S&L I had worked for survived the 80s -- without ever using FSLIC funds for losses . . . )

On to LTCM . . . I think you are right that at the center of LTCM was a (limited) partnership (of a group of principals). But technically, I believe LTCM was a hedge fund management group. Investors "deposited" money into the fund (without being partners), and received a return on their investments (you're right -- technically they were not stock dividends -- but I was heading toward a more general conception of "risk" vs. "social insurance" . . .). The pooled funds were then "managed" by LTCM. Hedge funds have an interesting status. Unlike the more familiar mutual funds many of us invest in, hedge funds are (as far as I know, still -- but certainly at the time of LTCM) essentially unregulated. They were (are) not under the usual purview of the SEC, did not have to follow disclosure rules, etc. In "exchange" for being unregulated, there were (are?) limitations on investors. For some such funds, the minimum investment was $1 million. For LTCM, I believe, the minimum was $10 million. ÊPart of the notion was that such funds were restricted to "savvy investors," who would understand and be able to bear the "risk" they were facing.

In general, I guess I agree that kurtosis of asset price deviations and "fat tails" (in general) had something to do with their problems. But, there were some more "traditional" issues -- leverage and liquidity. As I understand it, at the time of their troubles, they had leveraged an investor pool of something like $4 billion out to over $1 trillion in exposure (more than 250 to 1 -- and I seem to recall estimates that at times they were leveraged over 500 to 1! -- I think current rules for OTC derivatives brokers limit exposure to 30 to 1 or so . . .). Beyond that, much of their exposure was very much illiquid. Put those two together, and let financial markets have even slightly "fat tails," and you are ripe for a disaster, it seems to me . . .

An aside on how they got into the fix they were in (warning -- possibly unwarranted editorializing :-) : Among the principals in LCTM were Scholes (he of the Black-Scholes options pricing model) and Merton (watch for the possibly not coincidental link :-), one of whose "claims to fame" was (I believe) extending Black-Scholes to include the possibility that a stock being optioned paid dividends during the option period (!). They won the Nobel Prize in Economics in 1997 for "a new method to determine the value of derivatives." Do we anticipate, maybe, a slight case of hubris? They had this bright, shiny "new" formula/model for "controlling risk" that they (apparently) believed made them immune to the usual risks that investors faced . . . Anyway, it all came crashing down around them. (I read a moral here about trusting your models too much, and believing that your model is so "realistic" that it captures "all the essential characteristics" of the phenomenon in question. I like models and simulations, but I'm nervous about reifying them . . . I had some experience with the real-life application of models and computer simulations (and troubles they can get you into) in the early 70s at that S&L I worked at -- but that's another story . . .

When the crunch came (associated with a collapse in the Russian economy . . .), LTCM was poised to "go under." Eventually, Alan Greenspan (Chair of the Fed) stepped in and brokered a deal (with a group of banks) to "refinance" LTCM (in essence a new set of investments in the fund) along with restructuring. ÊGreenspan did not want there to be a "fire sale" of the LTCM assets, as had been done with the S&L assets. This was not a "bailout" in the same sense as the S&L bailout (commiting taxpayer dollars). However, it is worth thinking about why the refinancing banks were willing to invest in a failing enterprise. ÊThere were potential benefits to the new (and old) investors: for some, at least, protection of existing investments in LTCM; protection against a "domino" effect (i.e., general market stabilization, etc.); and perhaps others. Balanced against those potential benefits, however, were risks. Although the Fed (via Greenspan) did not directly commit taxpayer dollars, nonetheless one would presume that the banks considered the fact that the Fed was proactively working to "save" LTCM in judging the risk they were taking. But there are also larger range risk perception issues in play. As Greenspan himself said " Of course, any time that there is public involvement that softens the blow of private-sector losses--even as obliquely as in this episode--the issue of moral hazard arises." See Federal Reserve Board testimony.

The general "moral hazard" to which Greenspan refers is the perception among investors that Greenspan, by his actions, could be (and perhaps was) taken as commiting taxpayers as "lenders of last resort" even for unregulated high-risk investments like LTCM-style hedge funds. Greenspan's actions (although not directly commiting taxpayer dollars in this case) indicated that it might not be unreasonable for investors to believe that taxpayers were in effect sheilding them from some of the risk. In other words, one could (and Greenspan in effect did) argue that there was an element of "social insurance" in his actions. ÊGreenspan made a risk-balancing judgment (between just letting LTCM fall, versus intervening, and thus, indirectly, indicating that the taxpayers were bearing part of the risk), and determined that the risks to taxpayers inherent in letting LTCM fall were greater than the risks to taxpayers inherent (implicitly) in his active intervention. Are these risks hard to quantify? ÊYes. But that doesn't mean they are not there. In this case, one doesn't know exactly what Greenspan would have done if refinancing couldn't be arranged to shore up LTCM, but it is not unreasonable to imagine that he might have set about directly commiting taxpayer dollars. To put it a slightly different way, I can be insured against risk even if I never get a dollar from my insurance company . . .

So -- could it make sense for investors to pay a "premium" for enhanced protection against risk they enjoy by virtue of the existence of a (taxpayer) "lender of last resort?" Doesn't seem completely unreasonable to me.

What about corporations themselves? Do they enjoy some "social insurance" protection against risk? I'll just give one quick example. Bankruptcy laws provide legal protection for (at least some of) the assets of corporations. ÊAlthough we might say that the government "doesn't step in" and "lets corporations fail," consider an example like Worldcom. Right now Worldcom is "enjoying" the protections of bankruptcy proceedings. Could it make sense for corporations to pay a "premium" for protections against risk (obviously not complete protection against all risks, but some protection nonetheless) they have by virtue of things like bankruptcy laws?

There are various forms that some of these "social insurance" processes could take. Some, like FDIC or FSLIC, involve direct premiums for specific coverages. ÊOthers, like the S&L bailout, involve general taxpayer coverage for "rare or extreme" risks. Some, like LTCM, may be unusual (and in the specific case of LTCM did not involve direct taxpayer dollars), but can carry the message of implicit risk coverage. Others, like bankruptcy laws, are legal protections against certain risks, but without direct payment of dollars.

Another piece of "social insurance" processes is things like mandatory seat belt laws. There are costs we share socially from people being injured in accidents (e.g., some of the medical services provided to accident victims are not paid for by them). Thus, it is in the general social interest to have laws mandating seat belts. In this case, we end up paying a little more for our cars, but by doing so we are reducing risks -- rather than paying more in taxes to pay for social health services (i.e., covering financial risk for accident victims), we reduce the risk in advance (i.e., before accidents happen). One could argue that dividend taxes (to some extent, for some people) may serve to inhibit them from investing in the stock market. It may be a good thing for some people to avoid the risks inherent in stock purchases (I have heard estimates that as many as 80% of "private investors" playing the market end up losing money . . .). ÊPerhaps a little extra reminder in the form of dividend taxes serves a good social purpose . . .

Some of these issues raise some interesting dilemmas. One potential problem is that if people (investors or drivers?) perceive themselves to be experiencing reduced risk, they may actually engage in riskier behavior. For example, for a few years after mandatory seat belt laws, there was essentially no decrease in total injuries from accidents. People, feeling safer with their seat belts on, drove less carefully, and thus were involved in more accidents. Fortunately, over the years, things have improved . . .

A significant part of Greenspan's notion of "moral hazard" was that by intervening in LTCM, he was likely to be giving hedge fund managers the idea that part of their risk could be covered (e.g., by taxpayers), and hence they might engage in even riskier investment practices. This raises some interesting issues about how and when we (as a society?) might intentionally or inadvertently encourage or discourage certain forms of "risk-taking," and also some questions about how we might reduce risks in advance rather than, perhaps, helping pay for losses incurred from "risks" that went bad.

So anyway, back to dividend taxes for a moment. Reducing dividend taxes could have any number of possible effects. These could include: putting more money into the hands of consumers who would buy more (thus "stimulating the economy"); stabilizing stock prices; making stocks a more attractive investment and thus increasing stock purchases; increasing corporate investments in production capacity because it is easier to raise funds by issuing new stock offerings; decreasing corporate investment in production capacity because investors want corporate profits to be distributed as dividends instead of being used otherwise; decreasing private savings in bank accounts because individuals invest more in stocks; etc., etc. . . . Various of these are indirect and conjectural, and could be affected by a variety tax and other policy changes.

One trouble in all this is arguments of the following form: If we do X, then (the good thing) G is likely to happen. Therefore we should do X. A difficulty with this argument is that in practice it may be the case that we are more in the situation: If we do any one of X, Y, Z, then (the good thing) G is likley to happen. We might be able to conclude that we should therefore do one of X, Y, or Z, but not that we should do X in particular. Another difficulty with this form of argument is that it can hide other features of the situation, such as: If we do X, then (the bad thing) B is likely to happen (i.e., we may end up doing benefit analysis without doing cost analysis . . .).

So, how do we proceed? Obviously, in the end we have to just do the best we can given limited knowledge and limited predictive capability. Thus, we probably look for (what we perceive to be) dominant effects, and most likely effects. ÊBut even beyond that, the very judgment that G is a good thing, or that B is a bad thing, may depend on other assumptions and judgments. So, at least a few "judgment calls" ahead:

I think that cutting dividend tax rates has two dominant effects. First, certain investors will directly benefit from the cut (different investors in different amounts). Second, stock investments will become more attractive than before, which will encourage investors to buy more stocks.

Now we have a whole raft of evaluations we can make. Who benefits directly, and by how much? Do we evaluate this as averages, separated averages by categories, or number of individuals benefiting by given amounts, or by average amounts, or maximum versus minimum amounts, or etc., etc., etc. Who will be encouraged to buy more stocks, and do we want to encourage them? Will this encourage individuals to buy stocks instead of, say, putting money in savings accounts? ÊWill this encourage individuals who can't actually afford the risks associated with buying stocks to do so?

This leads also to evaluating alternative tax change policies. For example, might we be better off cutting income tax on corporations rather than cutting dividend taxes? Might other options be better? Here's one to consider: Êinstead of a tax cut on either dividends or corporate income tax, how about giving a bonus (in their IRAs) to people who put money in an IRA, and phasing in a (corresponding, by some formula?) reduction in social security benefits, thus helping to reduce the long term shortfall in the social security trust fund. ÊOne could structure this so that the net effect on (available) tax revenue is essentially the same as reducing the dividend tax (in the short term), and would directly increase individual savings and/or stock/bond investments (and thus make resources available for corporated investments in production capacity). ÊI'll leave it to you to think about other possible effects, both positive and negative, of such a change in tax policy. There are obviously many different kinds of things one can emphasize and affect in changing tax policies.

I realize this is getting (very :-) long, but the Orange County example is too interesting (to me :-) to pass up . . .

Let me start with (my view) of a couple of significant contributing factors to the Orange County bankruptcy.

In the "good old days," when local governments (city and county, for example) experienced increased demand for services (like police or fire protection, say), or increased costs for such services, there were relatively straightforward mechanisms to increase taxes to pay for the increases in services. Then, along came proposition 13, which significantly curtailed the possibility of such tax increases, and at the same time reduced tax revenues. However, even in strongly conservative areas such as Orange County, the desire for services did not abate. ÊThis led local governments to exploring other avenues for increasing revenues, such as investment portfolios. Local governments pooled their available funds, and had people (such as Robert Citron) act as "fund managers." These managers were consistently pushed hard to increase their return on investments. (Note that taxpayers in Orange County may have paid less in taxes, for a while, but ended up "paying" the costs of the loss of the investment funds. Might they have actually been better off just paying slightly higher taxes in the first place?)

A second significant contributing factor was that Citron started investing in Over-The-Counter (OTC) derivatives. These derivatives were (at the time) unregulated. They weren't stocks or bonds, so the SEC didn't regulate them. ÊThey weren't based on "commodities," so the CFTC and CEA didn't regulate them. ÊVarious brokerage firms (such as Merrill Lynch) started developing and selling their own derivatives. In effect, they began scooping off the "risk" from various "investments," bundling the "risk" up, and selling the resulting derivatives. These were packaged in such a way that they had high possible return, but they were very high risk. Robert Citron apparently didn't understand all the possible risks (he argues that he was misled by Merrill Lynch brokers, and there were lawsuits -- Merrill paid a settlement of over $400 million).

You may have noticed a recurrent theme in these examples -- reduced (S&L) or absent (LTCM and OTC derivatives) regulation. It is relatively easy to argue that an elected official like Robert Citron should have taken some finance courses, and thereby could have avoided getting into trouble as he did. But -- consider again the example of LTCM. The principals of LTCM included two Nobel prize winners in economics, who in fact won the Nobel prize precisely for their work in determining the "value of derivatives." And yet, they themselves went bankrupt, and (so we are told) nearly toppled the whole economy besides!

I would argue that sensible, fair, and carefully thought out tax policy, along with meaningful regulation are at the very least a reasonable exchange for the "social insurance" protections against risk available to investors within our economy. I would also argue that the current trend in tax policy, both at the State and National level, is putting our future at great risk. It looks to me as though we are going into a "short-long" mode, where we are piling up long-term debt in exchange for short-term "profits," and trusting that things will work out fine. I feel as though we (as a State and as a Nation) are doing more or less the same things with our economy as Robert Citron did with the Orange County funds -- trying to leverage short term "fixes" to cover long term risk exposure and debts.

Oh, well. I guess that's it for another one . . .

tom

p.s. I'm just finishing up my "taxes," which must explain why I'm so willing to write in defense of taxes! :-)