Remedial remittance arithmetic

As a follow-up to my rant earlier today, I thought I would write down what determines Fed profitability and thus remittances to the Treasury.  Maybe, I could try the tougher task of being correct, rather than just critical.  I have been whining a lot lately, you twelve may have noticed.

Bloomberg spreading nonsense still ticks me off, but it ain’t about Bloomberg, in fairness. I have developed an allergy to journalists getting simple stuff wrong in the past year. But it matters more in politics, than in finance.  So I should be clear that I pick on Bloomberg only because they cover stuff I would claim to understand and where normative issues are most easily bracketed.

Here I am just going to ignore two things that deliver a bit of slippage between Fed profitability and remittances. First, remittances need not be timed to the achievement of profits, which seems neither feature nor flaw, to me. Second, some might say that not all profits earned at the Fed are returned to Treasury, because the Fed pays “dividends” to private banks for their capital. I view that as a fixed cost.  In any case, it is small beer by the standards of the trend in Fed profits and the variation around it. So I ignore it, along with timing issues.

When the Fed has a normal balance sheet comprising only bills on the asset side and no interest-bearing reserves on the liability side, the source of Fed profitability is simply the gap between market interest rates and the zero cost of finance the Fed pays on currency (and much less importantly) reserves. Traditionally, this is called seniorage on “the” monetary base. But I am going to call it pure seniorage, to distinguish it from another source of profitability that has arisen recently, when the Fed has had a balance sheet that has been abnormal, at least by historical standards.  Importantly, the interest elasticity of demand for base money is low enough, i.e. close enough to zero, that pure seniorage correlates positively with rates, at least with rates in their plausible range and at horizons long enough to be relevant for budget analysis. People get that backwards.

We may describe the Fed’s balance sheet as “abnormal”, at least by convention, when the size of the asset side exceeds that required to meet currency and non-interest-bearing reserves demanded at the target fed funds rate.  Recently, the abnormal balance sheet has involved a large stock of interest bearing reserves (and other interest bearing liabilities, which we will set aside for simplicity) on the liability side and longer-maturity government-like debt on the asset side.  We call that QE.

In this environment, there is – or more strictly, may be — a second source of profitability, that arising from positive carry.  Lets’ call that trading profitability, because it involves a speculation on the path of short-term interest rates, which – in fairness – the Fed may be uniquely qualified to engage in.  (Hence the rates signal aspect of QE.)  And let’s be careful to distinguish it from what I am calling pure seniorage.

I think people, such as for example Stephen Stanley, sometimes overlook that the source of trading profitability is not the low-level of the fed funds rate or thus the low rate paid on excess reserves. And nor is it the positive slope of the yield curve.  To see this easily, just imagine a counterfactual (or recent experience) in which the slope of the curve less than fully compensates for a possible rise of short-term interest rates.  In such an environment, carry trades held to maturity of the asset leg would generate losses.

I guess how you analyze trading profitability is a matter of (modeling) taste.  But these three sources sum, by construction, to all of it. We can think of trading profitability as reflecting three things:

  • Luck
  • The Fed’s forecasting edge
  • The term premium in the yield curve

The first factor cannot be relied upon going forward. At some level this is painfully obvious. But I think some of those claiming that a lower balance sheet just necessarily means less profits / remittances may be forgetting this and just extrapolating the history of the past decade, during which carry trades – of the sort QE represents or, hell, is – have been very profitable.

The second factor is perhaps a bit more reliable, at least at short horizons, on the grounds that the Fed has a better insight into the path of the funds rate over the very near term than the market does.  But, the Fed demonstrably does not have a better view than the market at even medium term horizons, as the history of the past decade has shown clearly.

And whatever the Fed’s medium term advantage might have been, the escape from zero bound will predictably make it less compelling. In the interest of time and not losing the plot, I will leave figuring out why that is the case as an exercise for the reader.

So that leaves us with the third item.  The term premium is a very important determinant of the prospective profitability of carry traders, particularly among those – like myself – who still have some modicum of respect for the idea that the fixed income market might often be roughly efficient.

One problem is that term premium is not directly observable and there are good grounds for skepticism about attempts to estimate it.  Right now, the term premium does not seem to be very large, I would point out with appropriate humility.

But the point is that the Fed cannot control the term premium, even with QE, I would say. And for those who say that the Fed can control the term premium with QE, then even spotting that premise does not get you very far in this context. After all, asset rundown would raise the term premium, and thus lift the prospective profitability of the remaining carry trade, arising out of the remaining QE assets. For example, if cutting QE assets in half raises the term premium from 0 to 50 bps, that is a win for total remittances over time.

I have said it before and I will say it again. QE will have been conceived, implemented, scaled up, tapered and run off before the consensus even has the remotest clue of what QE is or how it affects things.  Getting its effect on Fed remittances wrong in just par for the course.  For me, that is a recurring theme.

Yeah, that’s the main point. Whatever your take, even if you insist on holding to the wrong one, all this remittance stuff is peanuts relative to the broader trends in the budget and egregious lying about them.  It is interesting much more as an example of how QE is intellectual kryptonite, turning even sensible people’s brains to mush.