E-piphany

It was another slow day, part of what is shaping up to be a typically slow August week.

I am always fascinated during these slow weeks by the fact that the same news that ordinarily would send markets spinning one way or the other will often seem ignored altogether, as if each hair-trigger trader is waiting for someone else to make the first move which never, as a result, occurs. Other times, a possibly less-significant item will trigger a bigger move if only a few large positions try to move through the illiquidity.

What that probably reflects is that very large traders – pension funds, large hedge funds, money managers – recognize that when liquidity is low there is a larger cost to initiating any move. Therefore, it takes more certainty of the result before it makes sense to re-allocate any meaningful amount of the portfolio. (Thus, the decline in volumes we have seen this year could be seen as deriving either from a lack of confidence about market direction, or from a decline in liquidity, or both.)

Today’s news was in the form of an interview in the Wall Street Journal, later followed up by an interview on CNBC, of Boston Federal Reserve Bank President Eric Rosengren. Now, Rosengren is a known hawk, but he called out his cohorts on the Fed to “launch an aggressive, open-ended bond buying program that the central bank would continue until economic growth picks up and unemployment starts falling again.”

This is monetary idiocy. Mr. Rosengren has just become the Krugman of monetarism: it isn’t working, thinks Rosengren, because a couple of trillion just isn’t enough to make a difference. I have renewed sympathy for Chairman Bernanke, if he is forced to deal with people like this who don’t understand what they’re doing, but figure they just need to do more of it.

Let’s be clear on the theory: if the Fed increases the money supply while the velocity of money remains static, nominal GDP  (PQ in the monetarist equation) will rise. But here’s where it’s important to actually understand the theory, rather than rely on an equation. Nominal GDP can grow for two different reasons: because the real economy has expanded (Q) or because the prices attached to all transactions has risen (P). Theory says that if economic actors are fully rational, they will recognize that the increase in money lowers the value of each transactional unit of money (dollar) and so the increase in M will be fully mirrored in P. If economic actors are at least somewhat stupid or naïve, and take the increase in the money in their bank account as an actual increase in wealth, they’ll spend more and the real economy will benefit.

This is called ‘money illusion’, and the evidence of the last few years is that it’s pretty weak. I suspect that’s because most people judge the balance in their checking account in two ways. First, they notice when the balance itself is increasing over time. But second, and significantly, they notice that each month the checks they write take more out of the balance than they previously did. That is, their reference point is not just the balance itself, but the interplay of balances and consumption. This makes it hard to fool them with money illusion. The Fed’s continued talk about how “inflation expectations are contained” is clearly partly intended to increase the money illusion effect and thereby increase the efficacy of monetary policy on the real economy – the ethics of that practice I will leave to others to discuss.

So if Rosengren had his way, and the Fed bought a trillion dollars of securities every month, it wouldn’t have a big effect on the real economy. But you can bet it would have a huge effect on the price level!

Now one place that I actually agree with Rosengren is on the interest paid on excess reserves (IOER). He said the Fed should reduce IOER, as I have written numerous times, and moreover that they should do it gradually so as to make sure it didn’t disrupt money market funds. Oddly, he said he didn’t want to go all the way to zero, so he’s arguing about maybe a 10-20bp ease, but since results to such a policy are likely to be non-linear it’s not unreasonable to go slowly.

Maybe it is talk like this that explains why inflation breakevens have recently been striking out higher. To be sure, another reason for the rise in inflation expectations, at least at the short end of the curve, is the 17% rise in spot gasoline prices since June 21^st, but this shouldn’t cause a severe effect at the 10-year point of the inflation curve. 10-year inflation expectations as measured by inflation swaps are up 25bps over the last two weeks, and breakevens (the spread between TIPS yields and Treasury yields) has risen by a similar amount.

This is an unusual time of year for breakeven inflation to be rising. As the chart below (Source: Enduring Investments) illustrates, compared to the last ten years’ worth of data on 10-year breakevens it seems almost as if this year’s pattern has been shifted earlier by about two months.

I don’t have a great explanation for this; most likely, it’s just spurious. But it helps to illustrate that this is an abnormal behavior. In the last 13 years, 10-year breakevens have declined in the 30 days following July 25^th on ten occasions, and this is also true (10 out of 13) at the 60-day horizon. The average additional “normal” decline in breakevens forward from this date, as you can see from the green line above, is about 15bps.

Now, that may mean that TIPS are overextended (relative to nominal bonds; there’s no question in my mind that they’re overextended on an absolute basis) and that breakevens are about to fall back. But it may also mean that there is something more significant happening here. I recently highlighted the unusual recent performance of commodities relative to the dollar, and this is of a piece with that observation. Our Fisher model has TIPS overextended, but also has inflation expectations lower than they ought to be, so that effectively it indicates a short position is warranted in both TIPS and nominal bonds rather than one versus the other (it first signaled this on July 31, for the record). The model signals go back to 2001, and this is the first time that we have ever had that configuration indicated.

Something interesting is happening, indeed, even if it is August.