Debt Wars Episode II: Attack of the Traders

My post from last week on debt generated some conversation about leverage and the value of arbitrage-free prices. Basically I'm arguing that if leverage (a.k.a. trading debt) is exceedingly expensive, then arbitragers won't be able to perform their essential function.

I wanted to illustrate this in a bit more detail by giving a admittedly stylized example. As you are reading this, bear in mind that I'm merely explaining why some degree of leverage is needed to produce efficient markets. This shouldn't be taken as an endorsement of any particular level of leverage, simply as an argument that Wall Street leverage is basically good.

Let's pretend we live in a world where there is a discrete amount of "real" money that is investable in the short-run. That is to say that all long-term investors have fully allocated their investments and any change in their demand for investments only occurs in the long-run. We can imagine this sort of like a world of dollar cost averagers who put some set amount of money aside for investment at the end of each year. However during the course of the year, the level of investable capital is constant.

Now let's assume we're at equilibrium and the market is perfectly efficient. Suddenly a new bond issue comes to market. Who is going to buy it? Given that investors don't have any uninvested assets, the only way an investor could buy this new asset is if they sell other assets. This can't work, of course, because if one investor sells another investor has to buy. There isn't any available net capital in the system.

Now many readers are thinking to themselves that this is a dumb example, since obviously my initial assumption doesn't hold. We see capital flowing back and forth all the time. Investable assets obviously aren't fixed.

Or are they? Consider who makes up real money investors. Individuals? They might have some assets sitting in their checking account which could in theory be invested, but its clearly limited. They have bills to pay with their liquid money. Corporations? Similar to individuals. Mutual funds? On any given day, they only have what they have.

Maybe we can't say that investable assets are literally fixed, but I argue that in the short run its damn close.

Furthermore, real money isn't going to react to relatively small arbitrage opportunities. Let's assume the yield curve is dead flat at 5%. Let's further assume that ABC Corp has bonds outstanding due in 2018 currently trading with a 6% yield. Now ABC wants to sell new bonds which will have a maturity in 2019. What should the yield be? With a flat yield curve, the yield should be extremely close to the 2018 bonds.

But if real money has limited excess capital, there will be a supply/demand imbalance here. Real money will have to be enticed to bring in new capital, and therefore the absolute yield will have to be large enough to do the enticing. Relative yield doesn't apply.

Put yourself in this situation. You wake up one morning comfortable with your investments when some poor bond salesman calls you up and asks about these new ABC Corp bonds. You admit that you have some money in your checking account, but are you going to tie up your liquidity for a 6% yield? Or 6.5%? Or 7%? Depends on your own situation. Might have to be 9% or 10%.

Now let's introduce the possibility of fast money. They also have limited capital, but we'll assume they also have access to leverage.

So back to ABC Corp. They look to sell new bonds. Real money is strapped and wouldn't be enticed unless the yield is 8%. But fast money can afford to go long new bonds vs. a short of the 2018 bonds with a yield differential that is much smaller. If fast money can get 10/1 leverage, a 1% differential earns them a 10% IRR. Or even better, fast money can speculate that once ABC Corp has sold their large bond issue, the supply/demand picture will improve. That is to say, if ABC Corp wants to sell $1 billion in bonds, real money can't take it down. But once the deal clears, fast money can sell the $1 billion bonds in smaller chunks at a tighter spread. Maybe fast money would therefore take the bonds at 6.25% and try to sell them off at 6%. That would be about a 2% price gain. At 10/1 leverage, that's a 20% IRR.

Why do we care about ABC Corp's cost of funding? Because the markets are supposed to be an arbiter of capital. If the cost of capital is distorted by technicals, the market can't function in this manner.

This problem is most glaring when the market is in panic mode. Last fall, real money almost universally wanted to sell at the same time that fast money had no access to capital. Efficient markets completely broke down.

Again, it isn't that unlimited leverage is a good thing. But without leverage, markets can't work.