And as long as I’m reading Kevin Drum’s blog, he had a piece about monetary policy yesterday that I’m still trying to figure out how to answer. It relates to a longstanding debate about whether monetary policy is sufficient to pull the economy out of the low-growth doldrums that it’s been in since the financial crisis. Fundamentally, I don’t believe it is.

To follow the argument, it’s probably a good idea to start here with Bill Gross’s piece that prompted the subsequent debate. The heart of Gross’s argument is: currently there is little return to risk, as evidenced by low absolute returns on risky assets – low yields on long-term government bonds, low yields on high-yield bonds, low earnings yield on stocks, you name it. And when there is little return to risk, there’s little return to the game of capital allocation. Asset prices may rise because the Fed is pumping money into the system, but capital is not being redeployed away from unproductive to more productive assets.

[P]rices on financial assets have soared and central banks have temporarily averted a debt deflation reminiscent of the Great Depression. Their near-zero-based interest rates and QEs that have lowered carry and risk premiums have stabilized real economies, but not returned them to old normal growth rates. History will likely record that these policies were necessary oxygen generators. But the misunderstood after effects of this chemotherapy may also one day find their way into economic annals or even accepted economic theory. Central banks – including today’s superquant, Kuroda, leading the Bank of Japan – seem to believe that higher and higher asset prices produced necessarily by more and more QE check writing will inevitably stimulate real economic growth via the spillover wealth effect into consumption and real investment. That theory requires challenge if only because it doesn’t seem to be working very well.

Why it might not be working is fairly clear at least to your author. Once yields, risk spreads, volatility or liquidity premiums get so low, there is less and less incentive to take risk. Granted, some investors may switch from fixed income assets to higher “yielding” stocks, or from domestic to global alternatives, but much of the investment universe is segmented by accounting, demographic or personal risk preferences and only marginal amounts of money appear to shift into what seem to most are slam dunk comparisons, such as Apple stock with a 3% dividend vs. Apple bonds at 1-2% yield levels. Because of historical and demographic asset market segmentation, then, the Fed and other central banks operative model is highly inefficient. Blood is being transfused into the system, but it lacks necessary oxygen.

Paul Krugman responds as follows:

When I read Gross and others, what I think is lurking underneath is a belief that capitalists are entitled to good returns on their capital, even if it’s just parked in safe assets. It’s about defending the privileges of the rentiers, who are assumed to be central to everything; the specific stories are just attempts to rationalize the unchanging goal.

Interest, classically (and I do mean classically, as in Mr. Keynes and the [sic]), is the reward for waiting: there’s supposedly a social function to interest because it rewards people for saving rather than spending. But right now we’re awash in excess savings with nowhere to go, and the marginal social value of a dollar of savings is negative. So real interest rates should be negative too, if they’re supposed to reflect social payoffs.

This really isn’t at all exotic — but obviously it’s a point wealth-owners don’t want to hear. Hence the constant agitation for monetary tightening.

To which Kevin Drum in turn responds:

Real investment yields aren’t low because the Fed is keeping treasury rates low. They’re low because the economy sucks and nobody has much confidence in the future. This lack of confidence keeps cash on the sidelines, and this in turn means there’s a huge supply of investment capital competing for a small number of good investment opportunities. When this happens, any project that’s even halfway promising can demand very low rates because investors are all clamoring to be let in. They bid each other down, which produces low yields even on risky investments.

I suppose Gross has some baroque explanation for why this is the fault of low Fed rates, but if he does, it’s nowhere to be found in his investment letter. The boring truth is that if we want higher investment yields, what we need is a stronger economy, one where the middle class is likely to thrive and consumption will increase robustly. In that economy, there are lots of great investment opportunities and yields will increase.

Allow me to offer a fourth perspective on this mess of an argument.

Let me start with Drum’s response. The problem with it is that, supposedly, the Fed has the power to engineer a growing economy – or, an economy that grows at its natural maximum rate without triggering accelerating inflation. If QE hasn’t achieved that (and it hasn’t) then the question is why not. Gross isn’t really making an argument for tightening. He’s making an argument that QE is failing to achieve its objectives. Drum appears to be agreeing to the degree that he refers to “real” investment yield being low, and posits lack of confidence “in the future” as the reason why the economy sucks – rather than, say, a lack of confidence that the Fed will engineer higher inflation.

Now let me move on to Krugman. Why would rentiers object to a rising stock and bond market? Anyone who already has money has done very, very well. Gross is complaining that people like him are having trouble doing their job – and their job is allocating capital. When yields on everything are low, that job doesn’t pay very well. But that’s a job; it’s not just collecting rents.

Now, on to Gross. His understanding of what the world’s central bankers are up to is that they are trying to engineer a recovery in the real economy by pumping up asset prices and counting on the wealth effect to boost investment and consumption. But he presents no evidence that this is what the world’s central bankers are trying to do – in fact, there’s plenty of anecdotal evidence that at least some central bankers are, just like Gross is, worried that this is what they’ve done, but all the evidence is that they were trying to stimulate borrowing and lending the old fashioned way.

So if they’re all wrong, what’s going on?

From where I sit, what’s going on is that, contrary to the hopes of the advocates of QE, the zero bound still matters, because the shape of the yield curve still matters.

Classically, the Fed controls the Fed Funds rate, and it controls it by engaging in open-market purchases designed to manipulate the effective rate down to their target. This rate is the rate at which institutions that can borrow at the Fed window will lend excess balances to other banks that need to augment their balances. And this rate is effectively the left-most point of the yield curve – the yield for overnight money on which the rest of the curve is based.

The various rounds of QE have involved engaging in open-market purchases across the yield curve because the effective Fed Funds rate can’t go below zero. Now, if you buy what the market monetarists are selling, the point of QE has not been to bring down rates at the long end (and thereby stimulate investment by making term-borrowing cheaper); rather, the point has been to raise nominal growth expectations, which in turn would cause rates to rise at the long end of the curve (i.e., to cause bond prices to fall).

How is that possible? How could buying pressure cause the price of something to go down? Well, think of it this way. Every buyer implies a seller. So the real question is where the clearing price is – and what the presence of a new buyer does to that clearing price. Could the presence of a new buyer stimulate selling? It depends on what information the presence of said buyer communicates to the rest of the market.

The Fed is not a speculative buyer. So its purchases do not tell the market that prices are too low. If the Fed buys $1 trillion of bonds, then investors know the Fed has $1 trillion of inventory to liquidate. And the Fed’s mandate is to liquidate that inventory when either inflation has picked up or unemployment has dropped or both – in other words, when nominal growth has increased. Which, in turn, should imply higher long rates for government bonds. Investors, in other words, should expect the Fed to sell in the future into a falling market, which should not make them sanguine about buying alongside a Fed that was bidding bond prices up and up. So, following this logic, Fed buying would not result in bond prices being bid up and up.

Bill Gross is looking at the market, though, and saying: that’s not what’s happening. It looks like prices keep going up, yields keep going down, and that’s the sign of a weak economy. Which, as everyone in this discussion agrees, it is!

But does that mean we need long rates to go up? Yes it does – but the Fed doesn’t control long rates directly; it controls the overnight rate. (Well, it doesn’t actually control that either, but close enough for government work.) What drives long rates is expectations about nominal growth. If the Fed ended QE, and the market monetarists are right, that would make long rates drop. Carry would get even worse. To make long rates go up, the Fed needs to raise expectations about future nominal growth change.

The market monetarists argue that the Fed can with relative ease engineer that shift in expectations simply by declaring a relevant target. Just say you’re going to buy stuff – any stuff – until nominal growth hits 5%, and nominal growth expectations will relatively quickly go to 5%, which in turn will encourage investment to take advantage of those expected returns, which in turn will make them a self-fulfilling prophecy.

I’ve never been entirely comfortable with that contention, and a major reason is that I’ve never understood the channel by which open-market transactions reliably drive changes in expectations. Basically, how do you convince the market that the Fed can and will credibly engineer its desired outcome in a specific timeframe.

The last part – the time frame – is crucial, because if the argument is simply that, yeah, eventually nominal growth rates will get to the target, but there’s wide disagreement about how long it will take, then it might make sense for many market participants to ride along with Fed buying in an inflating bubble – the opposite of what the market monetarists want them to do.

I’m a bit out on a limb with the foregoing, but that’s how I reconcile what Gross is observing with the theory behind the central bank’s actions.

So what’s the solution?

The right solution is for short rates to go negative. When a normal recession hits, the Fed drops short rates well below where long rates are, resulting in an upward-sloping yield curve. That both reflects expectations of recovery and creates the conditions for one. The most basic carry trade is banks borrowing short and lending long – borrowing from depositors and lending to real businesses – but every other carry trade is some more exotic version of the same thing. That’s what Bill Gross is looking for.

But banks can’t pay a negative interest rate on balances because otherwise people will stuff cash in mattresses.

If we didn’t have cash – if all money were electronic – this would not be a problem. Then short rates could go negative, which basically means you’d earn a negative yield on balances and would get paid to borrow money overnight. That would create immediate and obvious incentives to deploy capital in ways that generated a return, even a small one in nominal terms.

I note that Krugman also agrees that what we need is negative real rates. But if short-term inflation expectations are very low, the only way to get negative real rates is to have negative nominal overnight rates. And even if you could convince people that inflation will go up in the future, you can’t convince them that it’s higher right now than it actually is. So without the possibility of negative nominal short rates, short rates will always be too high in our situation. And carry will be too low.

Unfortunately, eliminating cash is not a short-term option – though I do think it’s something to work on, and I hope the Japanese are working on it right now. With their negative population growth rate, they should expect low to negative nominal growth rates as far as the eye can see. That shouldn’t mean immiseration – indeed, it should mean more real resources for the rising generation – but they are going to need to have negative nominal short-term rates if they don’t want persistent economic distortion.

The other lever we have is fiscal policy. I need to think a bit more about whether the information flow implied by a rise in Federal borrowing versus the information flow implied by an increase in buying by the Fed. I think it makes a big difference what activity is being financed. Market monetarists don’t believe there’s really any role for fiscal policy in macro-economic management – in part for the very good reason that they think the Fed can neutralize any fiscal stimulus if it doesn’t change its targets, and in part for the other very good reason that they assume government spending involves some dead loss relative to private allocations of capital. I think that there are arguments to be made on the other side of the ledger, however. But this post has gone on long enough.