As expected, the U.S. Federal Reserve made no overt policy changes after today's meeting. The statement was little changed, save for a slight, but meaningful tweak to its language around balance sheet normalization. That being said, this does help set the stage for how the monetary policy debate will evolve over the rest of the year.
Here are some thoughts.
The balance sheet horse has left the barn
By inserting the phrase "relatively soon" in the section about balance sheet normalization, the Fed is telegraphing that it will start this process in September. In a plan laid out last month, the central bank will start by reducing its reinvestments by $10 billion ($6 billion from Treasuries and $4 billion from mortgage bonds). The Fed will continue at that pace for three months until it is only reinvesting principal receipts amounts over $50 billion per month ($30 billion in Treasuries and $20 billion in MBS). In practice this means that after 15 months, the balance sheet will be declining by somewhere between $30 billion-$50 billion/month (since many months, it won't get $50 billion in principal payments and it isn't going to actively sell anything).
Will the Fed continue this balance sheet reduction no matter what?
Assuming Janet Yellen remained Chair, I think the answer is yes. I'd add that this says a lot about how the Fed is thinking about QE in the future. This "set it and forget it" strategy suggests the Fed doesn't believe in the "stock effect." The stock effect is the idea that just by holding a large amount of securities the Fed is influencing monetary conditions. For example, by leaving a smaller amount of bonds outstanding it is effectively reducing the total supply, which in turn should lower rates. Such an effect makes sense, but academic studies have struggled to prove that it actually exists.
I've argued for a long time that the primary effect of QE is through signaling, meaning that the best way for the Fed to commit to keeping rates low for a long time is to also be buying securities. Perhaps this is a sign that the Fed is souring on the idea of a stock effect and instead favoring a signaling effect. Indeed, if you believed the signaling effect was strong, you'd want to carry out an extremely slow exit to QE so as to have a small of a signal as possible.
If that's true, it has some consequences for what might happen when we next hit a recession. Assuming rates hit 0% again (which is likely), the Fed may try other means of signaling the same intent without tying itself up in long-term bonds.
But wouldn't the lack of reinvestment effectively increase the supply of bonds that need to find buyers?
Yes, but you need to put it into context. Average daily volume for Treasuries is $526 billion so far in 2017. Mortgage volume is $205 billion. Assuming roughly 20 trading days per month, that means that at its peak, Fed activity will amount to 0.34% of volume. It might matter, but won't matter much. Also, we'd assume that since the market knew the Fed was eventually going to unwind its balance sheet, some of this effect would have been priced in all along. Certainly given how much the Fed has telegraphed this unwind, it should be very well priced in by now.
Is a December rate hike more or less likely?
The market was pricing in 45% odds of a December rate hike prior to the release. It is little changed since, although that probability is up about 5% over the last few days. The Fed changed so little in its statement, there's no surprise fed funds futures aren't trading any differently.
I think it would take a lot for the Fed to back down from a December hike. If they did pause it would probably be because inflation data showed even more weakness. However, the Fed is clearly worried about financial market conditions. So as long as valuations remain frothy, there's a good chance the Fed is going to keep slowly hiking. Still, I think if employment remains reasonably strong and financial markets don't fall off, a December hike is pretty likely.
Why did the market rally?
We got a minor rally in Treasury bonds after this release, but after Tuesday's large Treasury selloff, I wouldn't read much into that. There were some that thought the Fed might actually start the balance sheet normalization process this month, so perhaps that it didn't happen had some effect.
Overall, expect the media to be laser focused on two Fed-related things in the coming months. First, is the Fed going to hike in December, and second, will we have a new Fed Chair? The former doesn't matter much for pretty much any security other than very short-term bonds. If the economy is still reasonably strong but the Fed doesn't hike in December for whatever reason, they'll be hiking shortly thereafter. The precise timing doesn't influence financial market valuations much over any time period other than a few days.
Obviously a change in the Chair would matter quite a bit, but we don't know exactly how. If there is a change, the leading candidate seems to be Gary Cohn, who doesn't have a known monetary policy philosophy. Gun to my head, I'd say he'll be more hawkish than Yellen, but that's honestly little more than an informed guess. If President Donald Trump is going to make a change at the Fed, he probably wouldn't make that announcement until November or so.
As far as trade ideas, I like bets on the Fed hiking in December. You have a bit of an opening to put the trade on right now, as you'll probably get some volatility around next Friday's payroll number. I'd short 5-ish year bonds as the way to express this. If you don't get a payoff after the jobs number, I'd cover and wait for another volatility point. The carry is too expensive to just stay short.