What Will the Fed Do When It's Finally Time to Tighten?
Housekeeping note: if you missed my comment on CPI from Friday, you can find it here. And if you missed my Bloomberg Radio interview with Carol Massar on Monday, don't worry! I will post it when Bloomberg makes it available on their site.
One of the busier sessions in recent memory (although still well short of 1bln shares traded on the NYSE, which was the standard not that long ago) resulted in a sharp rally in the equity market with the S&P +1.2% on the day.
The trigger for this holiday treat was the "progress" in the budget talks and what investors see as the increasing likelihood that the 'fiscal cliff' is averted. Be careful, however; whatever progress there was is fairly speculative, and I suspect we will see a bad news wiggle before all is resolved.
It is ironic, perhaps, that what is moving the process closer to resolution is the Republicans' sudden refusal to be steamrolled, and to instead try and play the game rather than try to negotiate as if both parties were trying to reach a fair resolution. I refer to the fact that Speaker Boehner has begun plans to start a separate legislative track in the House of Representatives by passing a bill that would keep the Bush tax cuts in place for most Americans; the bill would not avert the spending cuts that would take effect as part of the "fiscal cliff," but would keep the government from reaching more deeply into citizens' pockets on January 1st. It is, therefore, just exactly what the Republicans would want in these circumstances: spending cuts without tax increases (although fewer spending cuts than they would like).
The fact that this is a good play from the standpoint of the Republicans was immediately apparent from the fact that Democrats wasted no time in accusing Boehner of not negotiating in good faith with the President, and the President himself abruptly began to try and compromise slightly from his heretofore rigid position.
Of course, the Boener plan won't pass the Senate because it will produce exactly zero Democrat votes, and if it somehow passed by luck it would be vetoed by the President, so it has no chance to become law. However, by putting the Democrats in the position of having to vote against tax cuts, it greatly increases the chances that both parties might negotiate to something that all parties hate, and therefore passes with flying colors.
In the US system, by Constitutional writ all revenue bills have to start in the House of Representatives, so by the very nature of this process the Republicans, who dominate the House, hold the serve in this negotiation. Incredibly, this is the first time they've shown any desire to use that advantage to produce a bill that represents something closer to their views.
As noted above, equities reacted very well to the Republicans' show of spine. I'd noted several weeks back that I thought the Republicans had little incentive to negotiate, since going over the fiscal cliff represents smaller government and this may be the only opportunity that party has to get smaller government in the next few years. If this move persuades the Democrats of this fact, and the President moves to address the spending problem rather than just trying to soak the rich, then the fiscal cliff may be averted. It's really important in a negotiation, especially if a true compromise is to be reached, that your counterparty knows that you may walk away.
Personally, I think the odds are still against this happening before year-end, but some resolution fairly early in the new year is probably odds-on. However, with the debt ceiling also approaching, 2013 may well see more of these cliffhanger negotiations.
Bonds, interestingly, sold off. You would think that the prospect for a smaller deficit, even marginally, would help the Treasury market but in this case I think investors are reacting to the fact that if the fiscal cliff is averted, it lessens the chance of near-term recession and brings forward the day of reckoning for the Fed. Today, 10-year Treasury yields rose to 1.82%, which is near the highest level since early May, and 10-year real yields rose to -0.73%. Over the last five days, nominal yields have risen 16bps, and all of that has come from real yields. That is, inflation expectations have barely moved and 10-year breakevens remain at 2.50%. Ten-year inflation swaps are at 2.77%, and the important 1-year inflation, 1 year forward has risen to 2.23%.
So, whether the 'day of reckoning' for the Fed is near, or far...what do they do, when they've hit that point? And, more importantly, what does it do to the market?
Let's assume that we are at some point in the future and either the Unemployment Rate has dipped below 6.5%, the forward PCE inflation rate has risen above 2.5%, or inflation expectations have become "unanchored." The first thing that the Fed will do is to stop unlimited QE: the statement does not imply that they will immediately start trying to get out of the hole they are in, only that they will stop digging the hole. But suppose that inflation continues to tick up - since the evidence is that inflation is a process with momentum. What does the Fed do next? This is the real question. How quickly can the Fed react to adverse inflation outcomes?
The traditional option is that the Fed raises the overnight rate. The Fed announces this move, but the important part is what happens next: the Open Market Desk (aka 'the Desk') conducts reverse repos to decrease the supply of reserves, or sells securities outright if it wishes to make a more-permanent adjustment. This causes the price of reserves (also known as the overnight rate) to rise, and the Desk adjusts its activity so that the overnight rate floats near the target rate.
The problem is that this won't work right now. There are far too many reserves in circulation for the overnight interest rate to be increased by reverse repos or small securities sales. In fact, if it wasn't for the interest being paid on excess reserves, the overnight rate would certainly be zero, and might even be negative because the supply of reserves greatly outweighs the demand for reserves. They are called "excess" reserves for a reason - the bank doesn't need them, and will lend them overnight for pretty much any available rate.
So in order for the Fed to push the overnight rate higher, it must first soak up all of the excess reserves in the system - about $1.5 trillion at the moment - by selling bonds. Obviously, this is not something that can be done in the short-term.
But this misses the point a little bit anyway, because it isn't the rate that matters to monetary policy but the amount of transactional money (such as M2). The Fed can set the overnight rate at 1% by simply agreeing to pay 1% as interest on excess reserves (IOER). But that won't do anything at all to M2, because it won't change the amount of reserves in the system and doesn't change the money multiplier that relates the quantity of those reserves to M2.
So the short rate is dead. It isn't going to move for a very long time, unless the FOMC decides to help the banks out by paying a higher IOER. And if they do that, it's not going to affect inflation so it would just be a sweet present to the banks.
Okay, so perhaps the Fed can sell those long-dated securities and push long-term interest rates higher, slowing the housing market and the economy and squelching inflation, right? That's partly right: the Fed can sell those securities, and it can push long rates higher (although the Fed has oddly claimed that if it sold those bonds, interest rates wouldn't rise very much, which makes one wonder why they did it in the first place since presumably the opposite would also be true and buying them wouldn't push rates down), and that would slow growth. However, it wouldn't affect inflation, because inflation is not meaningfully affected by growth (I've discussed this ad nauseum in these articles; see partial arguments here, here, here, and here). But you don't have to believe all of the evidence on that point; just play it in reverse: if driving long rates down didn't cause a sudden jump in inflation, why would driving long rates up cause a sudden dampening in inflation?
Fama, in that article I quoted last week, had a very good point which I thought it was worth developing in more detail. The Fed has its hands off the wheel with respect to inflation...which isn't a problem, except that they're sitting in the back seat. The back seat of a very, very long bus.
In any event the issue isn't when the Fed starts its tightening, but when inflation stops going up. These are not the same things. If core inflation were to start ticking higher today, at a mere 1% per year, I think it would take 6-9 months for the Fed to stop QE (core PCE is at 1.6%), probably another 3 months at a minimum before they started to tighten, and then at least 1-2 years before they could have any meaningful impact on the money supply and cause inflation to slow. Maybe I'm being pessimistic, or maybe I'm being a bit generous by assuming that after a year the FOMC would start doing something very dramatic to sop up reserves, like issuing a trillion dollars in Fed Bills, but even assuming that everything works out just about as well as it conceivably can, if inflation started heading higher in that way then you're looking at a core CPI figure of 4-5% before it stops rising. Like I said, it's quite a long bus, and that translates to long "tails" of inflation outcomes.
How would markets react to this? Obviously, bond rates would be much higher, but would this be good or bad for equities? The conventional wisdom holds that equities are good hedges for inflation, because over a long period of time corporate earnings should broadly keep pace with inflation. While that is true, it is also the case that earnings tend to be translated into prices at lower multiples when inflation is high (a fact that has been known for a long time; in 1979 Franco Modigliani and Richard Cohn described this as an error but there isn't consensus on that issue) so that stocks tend to do relatively poorly when inflation is rising and better when inflation is falling from a high level. Moreover, stocks do especially poorly in the early stages of inflation when short-term inflation is surprising to the upside, as the chart below (Source: Enduring Investments) illustrates.
This chart highlights headline inflation, rather than core, but the point should be clear: nominal bonds and equities produce good real returns when inflation is surprising to the low side (even if that means that inflation is just going up slower than expected), and very poorly when inflation surprises to the high side (even when the overall level is low).
In my mind, this means that every investor needs to have some inflation protection, but especially now when the chances for an ugly inflation surprise are significant. For the record, the best asset class when inflation is surprising to the high side as measured here? Even inflation-linked bonds have produced negative real returns in such circumstances, because the real yield increase outweighs the higher inflation accruals in the short run. But commodities indices historically produced a 4% real return over that time period when inflation surprised at least 2.5% to the upside.
 It isn't clear to me why you would want to wait until they were unanchored, if anchoring matters, since presumably it isn't easy to anchor them again. After all, the whole reason the Fed wants anchored inflation expectations is because a regime change is thought to be hard - so if they are unanchored, you've just made it really hard to get inflation back down. In any event there's not much evidence that "anchored" inflation expectations matter to actual inflation outcomes, but it's just weird to me that the Fed would imply that they'd wait until expectations get loose from the anchor.