Abstract finance: Structured Investment Vehicles and other Tricks

Once you have securitization under your belt, it’s time to move on to SIV’s, structured investment vehicles. The basic idea is that one way to securitize financial value – whether they be assets, leases, or some other future income stream – is to dump them into a trust created for that purpose.

A trust is just an entity created to take in assets on one end, and dish them out to some beneficiary on the other end. The trustee’s job is to dish out those assets within the confines of the trust’s terms. A common way to think about trusts are pools of money set aside for rich kids so that they don’t get $1M and spend it all at once. But a trust can be much more flexible than this.

For instance, a trust can take in assets from various streams of financial value and then distribute these resources according any number of guidelines, in equal shares. Something like this (a ‘unit trust’) looks a lot like a mutual fund, in that a diverse amount of capital gets transformed into investor units. There may be rules about which assets get distributed first, second, third, etc., and to whom.

Financial services companies and hedge funds use trusts to do three kinds of things: 1) mix and match differentially-risked assets to ‘create’ an asset that conforms to a desired level of risk and return; 2) make investment assets opaque, or, in the words of Grégoire Mallard, to close the ‘black box’ of finance; and 3) ‘streamline’ (or avoid) regulation around foreign ownership of US assets, and other slightly-shady but probably not outright illegal tax dodges. This last use is probably the most important one in most instances. For example, additional tax burdens are in place for foreign owners of US assets, and if 1 in 100 investors is a non-US citizen, then an investment fund would have to mess with the tax system. A trust is often used to get around these rules. It is a case of de jure regulation in place, and burdensome, dodgy ways to get around that regulation (which is not as terrible a way to do things as it appears, but that’s a different post).

A specific kind of investment trust is – or was – a structured investment vehicle. Structured investment vehicles existed as a kind of fund from its inception by Citibank in 1988 until the last SIV was liquidated in October 2008. They are a kind of investment trust that sold short-term commercial paper, and then used the money to purchase longer-term bonds. In effect, they would borrow money in the short term, and then lend money in the long term. Since long-term loans had generally higher interest rates than shorter-term loans, they would profit on the difference.

Sound innocuous, no? Look at the yield curve for US Treasuries. Borrow in the short term, then keep rolling over your investments (borrowing again and again) as your long-term investments mature. This is what banks do, after all. However, SIV’s were designed not as banks, really, but as super-duper investment vehicles. About three-quarters of their long-term assets were not US Treasuries, but Asset Backed Securities (ABS), and Mortgage-Backed Securities in particular (MBS). These are assets based on bundled, securitized housing loans. These bonds were much higher-yielding than treasuries.

But even before these various asset-backed securities went boom, investment vehicles that take in assets on the one side and dish out distributions on the other side had been used to, well, screw investors for some time. Frank Partnoy noted a particularly nasty variety (in F.I.A.S.C.O., natch), known as the MX. In this tasty vehicle, zero coupons (bond payments with no interest attached to them, just a single lump sum) were combined with IOettes (interest-only slivers, the pieces of mortgaged backed commodities with high interest, high volatility, and robust-and-risky chances of repayment). Effectively, this was combining lead (the zeros) with gold (the IOettes).

So, the interest-piece was worth a ton, and the zeroes worth little. Then the two were smooshed together in a trust vehicle set up in the Caribbean. The ‘average’ cost of these instruments was booked by Japanese investors as a cost (gold + lead) / 2. Got it so far? Two pieces, one more valuable than the other, with their costs averaged.

And then, days later, half the units were sold. But because the investment vehicle was set up with its own particular rules for distributions, it required that the most valuable assets (the IOettes) were distributed in the first half of the trust units. The Japanese investors then booked the value of these units as revenue. Ta-da! In two days, profits galore! Of course, 29 years later the rest of the units (the lead) would be liquidated, at a massive loss. But by then who would remember? Who would care?

And this is the point. The abstraction of finance in theory provides an unambiguous social good – it efficiently distributes capital where it is most needed in the economy. In practice structured finance has been the playground for massive amounts of fraud (legal and illegal) for decades. Evidence for this predates by decades the craziness of 2007-

Comments are disabled for this post