In the movie Inside Job, George Soros makes an analogy that made an impression on me. He talks about how oil tankers have partitions in their hulls with their oil divided across the compartments. That way when the seas are rough the oil sloshes around within its own restricted space rather than the entire cargo splashing forward and back the full length of the ship, as would happen if there were no partitions.  This obviously makes the tanker more stable.

The analogy is to financial markets and regulation. Erecting partitions to make the market less liquid should improve stability.  At first you say, oh that’s a nice piece of rhetoric but financial markets aren’t anything like oil tankers.  At least I said that.

But let’s take it for a spin. Let’s analyze the partitioned tanker but forget that it is a mechanical system and instead analyze it in the same way we would use equilibrium theory to analyze a market.  Here goes.

There is a shock (rough seas) and the oil starts to spill to one end of the boat. But the partition stops the oil from going where it wants to go. There is a friction in the market. The oil in one compartment and the empty space in the adjacent compartment want to make a voluntary exchange.  And it would be Pareto improving (otherwise they wouldn’t want to do it.)  But that partition is stopping them.  This is welfare-reducing.

Moreover, there is a powerful incentive for arbitrage. Any small leak in the partition would allow equilibrium to be reached by removing the friction, allowing the oil to go where it wants to go, and relieving some internal pressure.  That must be welfare-improving.

If you think about it, market models pretty much stop there. Pareto improving trades should and do happen. Financial innovation brings down those partitions and that’s good. What is almost always missing is any way of talking about the hard-to-define but clearly very real externality that is the effect of these trades on the stability of the system as a whole.  That’s about process and transitional dynamics, not about equilibrium.

Indeed, in equilibrium the oil in the tanker is in the same place whether or not the partitions are there.

(drawing:  The Bigger Picture from