Federal Reserve Board Chair Janet Yellen said Wednesday that the central bank plans to keep interest rates at a record low "for a considerable period." (Reuters/Gary Cameron )
(Reuters/Gary Cameron )

Two years and $1.7 trillion later, the Fed's latest round of bond-buying, or QE3, is officially over. What did it get us?

Well, the best answer is what it didn't get us: a recession in 2013. Now the words "fiscal cliff," "sequester," and "debt ceiling" might be hazy memories from a time when Congress was doing its most to sabotage the recovery, so here's a refresher. Between higher income taxes on the rich, higher payroll taxes on the rest, and significant spending cuts, there's been an awful lot of austerity the last few years. Enough that the economy should have slowed down quite a bit. Indeed, the nonpartisan Congressional Budget Office estimated that all this fiscal tightening would subtract something like 1.5 percentage points from growth through the end of this year.

But that's not what happened. Consider this: the economy added 186,000 jobs a month in 2012, before all this austerity, but then accelerated, ever-so-slightly, to 194,000 in 2013 and 224,000 so far in 2014. Now there are plenty of caveats here. Just because the economy added more jobs with these cuts doesn't mean the cuts didn't hurt. We probably would have added even more if not for them. And besides, we would expect growth to eventually pick up on its own as households pay back what they owe. But all that said, it's still remarkable that the economy didn't slow down at all despite all the fiscal tightening. And for that, we can thank the Federal Reserve.

"The problem with quantitative easing," former Fed Chair Ben Bernanke has said, "is it works in practice, but it doesn't work in theory." That's never been more true than with QE3. Unemployment has fallen from 7.8 percent when it started to 5.9 percent now that it's over. And even though inflation is still below target, at least it hasn't fallen so much, like it has in Europe, that it's hurting the recovery.

The problem, though, is it's hard to say much more than this. We can look at how much long-term interest rates fell when the Fed started buying bonds, but that doesn't give us the full picture of how it worked. That's why it's worth going back and thinking about the theory for a minute. That'll tell us more about what QE3 did, and why the Fed shouldn't have ended it this soon.

Now quantitative easing is a lot simpler than it sounds. The Fed just creates money, and uses it to buy bonds from banks. But since the banks haven't lent any of it out, this all shows up simply as increased bank reserves. So why should this boost the economy? Well, it shouldn't really. The Fed has made banks trade bonds that pay interest for reserves that pay interest. That's a negligible difference. But it does work because banks end up trading super-safe bonds (which they sell to the Fed) for riskier ones. That, in turn, should push asset prices up, and make people feel rich enough to start spending more.

But there's another story about how QE works, and that's all about borrowing costs. Bond prices, you see, go up when the Fed buys them, and that pushes interest rates down. It's all very mechanical. Now that's not wrong, but it's not the full story either. Rates have stayed lower than you'd think considering that, at most, the Fed was buying $85 billion of bonds a month. That's because QE isn't just about purchases. It's about the promises those purchases imply.

It's confusing. But think about it this way: The Fed isn't going to raise rates as long as it's buying bonds. QE, in other words, is the Fed's way of printing its money where its mouth is when it says rates will stay low for a long time. That's why, as economist Michael Woodford argued, QE works better when it's used with forward guidance that makes the Fed's promises about future policy more explicit.

The question, then, is what message the Fed is sending now. Actually, we know the answer. The fact that it's ending QE3 despite still-low inflation and still-high, though declining, unemployment, signals that the Fed is eager to return to normalcy. So does changing its statement from saying there's a "significant underutilization of labor resources" to it "gradually diminishing." The Fed, in short,  looks much more hawkish. And that's not good, because, as Chicago Fed President Charles Evans explains, the "biggest risk" to the economy right now is that the Fed raises rates too soon.

QE isn't magic — far from it — but it is, as Boston Fed President Eric Rosengren told me, "quite effective." Especially at convincing markets that the Fed won't raise rates for awhile, which is all it should be saying right now. Because the only thing worse than having to do QE is having to do QE again. The Fed, in other words, should do everything it can to make sure the economy lifts off from its zero interest rate trap before it pulls anything back.

Otherwise, we might find ourselves back in the same place a few years from now.