telescopeRichard Feynman once remarked that “I believe that a scientist looking at nonscientific problems is just as dumb as the next guy.” I usually try to add that, as sometimes the next guy is me, that’s really remarkably dumb. However, I also try to widen it to the idea that entirely competent and sensible people looking at problems outside their own knowledge base end up looking dumb.

Not because they are indeed dumb or incompetent. Not on the level of my knowledge about programming for example, I don’t even know which way round you put the 1s and 0s into the coding machines. Rather, that in most fields there are some known but often unspoken assumptions, tricks of the trade perhaps, that you might not be aware of.

In my own specialty, that of weird metals, I get very tired of people telling me that mineral reserves are going to run out in 30 years. Reserves are simply the working capital of currently extant mines. It’s like complaining that GM is going to stop making cars in six weeks’ time because that’s all the steel they have in the warehouse. There’s simply no relationship between mineral reserves and the amount of any mineral that we have available to us, none at all.

Which brings us to the delightful subject of econophysics.

The physicists have noted that economics could do with a boost, is full of difficult math and is often describing chaotic systems. Physicists can do hard math, are willing to offer a boost and, as anyone who has ever watched the organisation of a round of drinks at a physics symposium will know, are well accustomed to chaos. So, perfect, let’s bring the insights of this hard science to one of the social ones.

An example is this paper, described at Bloomberg:

In a new paper, complex-system specialists Sebastian Poledna and Stefan Thurner offer a relatively simple idea: Make banks recognize systemic risk by charging them for creating it. Currently, lenders worry only about the risk they face from a default by the borrower, as opposed to the risk the loan may present to the larger financial system by making cascades of default more likely. Poledna and Thurner suggest that a small tax, proportionate to the systemic risk a loan entails, could help rebalance the incentives.

Woo Hoo! Great: no, really, great! A solution to systemic risk, that’s something that’s really going to make the economics world sit up and take note. The full paper is here and what they’re really describing is a form of Pigou Tax.

If by your action you increase systemic risk then you should be charged an amount that reflects that increase in systemic risk arising from your action. This is exactly the same as the justification behind the carbon tax to deal with climate change and it’s a sound and well understood technique. That it arises from a study of an excess of rabbits in one field upon the crops of the farmer next door (no, really) back in the 1920s doesn’t make it any less valuable as an insight or a prescription.

So I did a quick search of the paper to see where they linked their idea to Pigou and couldn’t find it: this led to a rather closer examination of what they are proposing. For if you’re suggesting something that’s a well known economic technique but you’re not making the connection between your proposal and that well known economic technique we might end up finding that your knowledge of the field under discussion could be, well, possibly a little lacking?

And sad to say we do indeed find this. The mistake is in this:

It is important to note that SR spreads by lending. If a
systemically risky node lends to a systemically non-risky
one, the later inherits SR from the risky node, since if
the non-risky borrower should (for whatever reason) not
repay the loan, the risky node would trigger systemic
damage. In this sense SR spreads from the risky through
lending.

Ah, no, that’s not what we worry about as far as systemic risk is concerned in financial markets. In fact, we worry about it entirely the other way around. It’s not at all about who the money is lent to: it’s about who has the money been taken from before it is lent.

Imagine the following: we’re going to invest $100 million in PJ O’Rourke’s mythical fur bearing trout farming operation. If the money comes from a VC fund then the inevitable failure has no systemic risk at all. Meh, so $100 large in equity got swallowed. If we issued bonds to finance this then we’ll have a rather larger problem and if we were using the supposedly risk free funds from fractional reserve banking then we might indeed find ourselves with said systemic risk. For the depositors of the bank that had so wasted their money on the project would descend in droves to get the remnants of their savings out. And given that a bank only ever has 2 or 3% of deposits on hand in cash we get a bank run and the collapse of the bank. Even if it was still solvent, it would be illiquid and thus bust as a result of people hearing about the failure of the investment.

That is, systemic risk comes not from who the money was lent to but who it was acquired from, on what terms, before it was lent. So our physicists are looking at this problem through the wrong end of the telescope.

The other problem with the proposal is that the economists have already got this one right. Both Obama’s bank tax proposals and the bank levy in my native UK deal with exactly this problem. It is the banks’ liabilities (in banking, meaning where they get their funding from, not who they lend to) which are the potential problem. Thus a bank that funds its assets through equity poses no systemic risk and would pay none of either of these taxes. A bank that funds through long term bond issues would pay a small tax and one that funded through the highly liquid short term money markets would pay a larger. For that is what that systemic risk is and where it lies: that the bank cannot pay back its funding rather than that the loan itself goes sour. No one particularly cares that Citibank doesn’t get its money back from the trout farm: everyone cares if Citibank starts defaulting on loans to Citibank.

I have a feeling that this problem is widespread: people commenting outside of their knowledge base often end up missing something important about the new field they want to comment upon. In this case the proposal looks entirely sensible if one doesn’t know the details of what constitutes systemic risk nor what is already being done to combat it. And why should a physicist know the details of either? But the end result is that they do come up a day late and a dollar short.

One day I’ll get to Jaron Lanier and his “Forward to the Middle Ages” plan which is a perfect example of just this problem.