The 2% Mystery: Why Has QE3 Been Such a Bust?

Ben Bernanke listened to his critics, but the recovery is still stuck. What went wrong?

(Reuters)

To print or not to print? That is the question dividing the Federal Reserve.

Back in September the Fed launched its latest, and most ambitious, bond-buying program to date, dubbed QE3. Unlike before, the Fed hasn't committed to buying a specific dollar amount of bonds with QE3; instead, it's committed to buying $85 billion of bonds a month until the labor market improves "substantially". But what's "substantial" and what's not? And what if the Fed loses its nerve before the economy arrives at this mysterious moment of "substantial" improvement?

This latter question has gripped markets after the Fed's January meeting when "a number" of members said it should "taper" its bond purchases even before, you guessed it, there's any substantial improvement in unemployment. In other words, an increasing, and increasingly vocal, minority at the Fed are nervous about keeping open-ended bond-buying quite so open-ended. Now, a vocal minority is still a minority -- and besides, Bernanke tends to get his way -- but this hawkish talk has been enough to spook markets that thought QE3 wouldn't end much before 2014.

But there's a better question than how long QE3 will last. That's how much QE3 will work. Let's back up for a minute. Whether you want to call it "quantitative easing" (QE) or "bond-buying" or "large-scale asset purchases" (LSAP), the idea here is fairly simple: the Fed is printing money and buying pieces of paper. It's doing this because it can't boost the economy like it normally does by cutting short-term interest rates; those rates are stuck at zero, and can't go lower. Okay, that's not entirely true. The Fed can't cut nominal rates now, but it can cut real ones -- in other words, it can push up inflation, thereby reducing inflation-adjusted borrowing costs. That's what the Fed has done by printing money and buying long-term bonds from banks. Even if this freshly-printed money ends up as bank reserves (which it mostly has), the Fed is signaling that it wants more inflation.

Right?

Take a look at the chart below of what markets (roughly) think will happen with inflation over the next 5 years, annotated with the Fed's unconventional policies. Markets expect more inflation every time the Fed eases, and less every time it stops ... until QE3. Then, almost nothing. That's crazy. QE3 is open-ended, whereas previous rounds were not. This difference should have convinced markets that this time the Fed was really serious about jump-starting the recovery. Has QE hit a wall of diminishing returns? (Note: The black line shows the Fed's 2 percent inflation target).

Look again, but this time, focus on the black line. QE has hit a wall, but it's a wall of incredibly well-anchored inflation expectations, not diminishing returns. In other words, the Fed has quite easily been able to push inflation expectations back up to its 2 percent target, but no more. QE1 and QE2 had big effects, because they came when expected inflation was well below 2 percent and falling; QE3 has not, because expected inflation was already around 2 percent.

But wait. The Fed unveiled the Evans rule back in December, telling us it wouldn't raise rates before unemployment falls to 6.5 percent or inflation rises to 2.5 percent. In other words, isn't the Fed's 2 percent inflation target really a 2.5 percent inflation target now? Not exactly. The Fed is telling us it will tolerate 2.5 percent inflation, not that it will create it -- indeed, the Fed doesn't think inflation will stray at all above 2 percent over the next few years.

The best way to figure out what the Fed wants is to listen. After all, it tells us what it thinks will happen with GDP, unemployment, and inflation over 1, 2, and 3-year periods. Now, it's GDP and unemployment predictions have been, in the spirit of generosity, a tad optimistic, but not so for inflation (which, not-so-coincidentally, is the only above variable the Fed controls directly). The chart below looks at the Fed's core PCE inflation projections since late 2008; upper-range estimates for 1, 2, and 3-year periods are in red, and lower ranges ones are in blue. This is what a 2 percent inflation ceiling looks like.

There's a lot going on here, but there's a depressingly simple message in this chart: QE3 isn't working, because the Fed doesn't want it to work. The Fed revised its inflation projections up after QE1 and QE2, and markets followed; the Fed has kept its inflation projections steady after QE3, and, again, markets have followed. Now, this doesn't mean QE3 is entirely useless -- it's at least stopping inflation expectations from falling -- just that it could be doing much more if the Fed let it. That would be simple enough. The Fed could make its forecasts symmetrical around 2 percent, rather than peaking at 2 percent. Or it could say it expected (or is that wanted?) inflation well above 2 percent over the next two years, but not after that; in other words, make its target more explicitly flexible.

That leaves us with one last question. The Fed has shown time and again it can push inflation expectations (which largely determine inflation) up to 2 percent, even when short-term rates are parked at zero. But is that as much inflation as the Fed can create? It's hard to see why that would be the case, other than the Fed's self-imposed 2 percent ceiling. But the great thing about self-imposed problems is you can stop imposing them. The Fed doesn't need a new mandate (like NGDP targeting) to speed up the recovery; it just needs to tell us it wants -- gasp! -- 3 percent inflation for a year or two.

Until then, the recovery will suffer the outrageous slings and arrows of our 2 percent ceiling.