Suppose the Treasury issues $100 billion worth of 3-month T-bills yielding approximately zero (as 3-month T-bills do today and likely will continue to do until some time in the unforeseeable future when the Fed raises its target rate). Does it make any material difference to anyone whether those T-bills are bought up by the Fed (i.e. monetized) or remain with the public?
First of all, does it make any difference to the public? To put that a little differently, does anyone care whether they personally are holding T-bills or cash? More precisely, not really anyone. After the Treasury sells $100 billion worth of T-bills (assuming that the Fed doesn’t buy them), there will still be millions of people who didn’t choose to buy those bills. Those people don’t matter: they obviously don’t care if the Treasury sells the bills to the public, because they won’t buy them either way. They might care about the possible economic and financial effects of monetization, but, as I will argue, there aren’t any effects to care about.
So let’s look at those people (let’s call them people, even though IRL they’re mostly institutions) who are currently holding money and who will buy up the $100 billion worth of T-bills if the Fed doesn’t do so. Does the Fed’s action or lack of action make any difference to those people? Obviously it must make at least a tiny bit of difference, or they wouldn’t have bothered to buy the T-bills.
But it makes only a tiny bit of difference. Money yields zero; T-bills yield zero. Money is slightly more liquid than T-bills. But only ever-so-slightly: the market for T-bills is extremely efficient, and the price variation is minimal (especially given the Fed’s policy of only changing its target in quarter-point increments). There is only the tiniest risk of being unable to sell a 3-month T-bill almost immediately at any time at a price close to the price you paid for it. Aside from liquidity, T-bills are slightly safer then money. But only ever-so-slightly: if you’re an individual, you can distribute your money across banks and have it 100% FDIC insured; if you’re a bank, you can hold deposits at the Fed, which are possibly even safer than T-bills. It makes no material difference in which form you hold your assets.
But if the monetization doesn’t make any difference to the public, does it make a difference to the Fed or the Treasury? Let’s take the Fed first. The Fed can create and destroy money at will. The Fed will be able choose, with no constraint or cost either way, whether to roll over the T-bills when they mature. Moreover, like the public, the Fed can sell the T-bills, very quickly and with little price risk, before they mature, if it should decide to do so. So the only way it would make a difference to the Fed is if the purchase of T-bills has some economic effect that the Fed cares about. But, again, as I will argue – as I am arguing – there are no economic effects.
What about the Treasury, the government? Surely the government cares whether it really owes money to someone out there in the world vs. merely nominally owing it to the Fed. Actually, no. As noted above, the Fed can create and destroy money at will. If the Fed does buy the T-bills initially, it will still be able to choose whether or not to roll over the T-bills when they mature, and it will be able to choose whether to sell the T-bills before they mature (in which case the Treasury would subsequently owe money to the public again). Unless (as I again deny) the monetization has some economic effect, the Fed will continue to be indifferent, as long as the conditions of my initial assumption hold (i.e. until the T-bill yield rises above zero, which would have to be the result of a choice by the Fed to raise its interest rate target). And since the yield is zero, the Treasury pays no interest on the T-bills either way.
Suppose we do get to the point where the Fed raises its target rate. First take the case where the Fed had not monetized the debt initially. Suppose, for example, that, to get the target rate up, the Fed has to sell $200 billion worth of T-bills. Fine. Now take the case where the Fed had monetized the debt. In that case, the Fed will now have to sell $300 billion worth of T-bills. After the transaction takes place, the Fed’s balance sheet, and everyone else’s balance sheet, will look exactly the same in one case as it did in the other. The only difference is in what those balance sheets looked like before the Fed decided to raise the interest rate. And that difference, as I have argued, is inconsequential to all the parties involved.
Except of course if it has some economic effect. But the only way it could have an economic effect is if it changes someone’s behavior. And, since it has no material consequence for anyone, it won’t change anyone’s behavior.
Well, OK, it might. The only way it might change someone’s behavior is if they expect it to have an economic effect. Then the existence of such an effect would become a self-fulfilling prophecy. That’s what economists call a “sunspot” (by the analogy that literal sunspots will have economic effects if and only if people expect such effects). I would suggest that, even in that case, the effect is likely to be quite small. If there is no fundamental reason to expect an economic effect, there should be plenty of people speculating against those who do expect an effect. Moreover, if there is no fundamental reason to expect an effect, while one can still imagine that someone might expect some effect, it’s hard to see how anyone could expect a large effect, unless their reasoning process is seriously screwed up (in which case they aren’t likely to have much wealth left to allocate). With some people expecting not-too-large effects and other people speculating against them, it’s hard to see how the net impact on markets could be significantly large.
Now you might say, so much for your example of short-term T-bills, but the subject of this essay was whether or not to monetize, and the Fed has been talking about the possibility of monetizing long-term Treasury debt as well. Won’t that have an effect?
But again the answer is no – as long as the Treasury is flexible enough to choose its preferred financing option in either case. How much of Treasury borrowing will be long-term and how much will be short-term? That is entirely the Treasury’s decision. Suppose the Fed decides to monetize long-term debt instead of short-term debt. If the Treasury’s preferences are unchanged, it will simply issue more long-term debt and less short-term debt, and there will be no difference in the quantity of each type of debt held by the public. The only difference will be what is held by the Fed. But that is no difference at all, since the Fed’s profits go directly into the Treasury. It is as if the Treasury owed the money to itself. Why should the Treasury care whether the money it owes to itself is booked as a long-term debt or a short-term debt? Moreover, since the Fed can buy and sell any amount at will at any time in the future, the Fed, counting on the Treasury’s indifference, should also be indifferent.
But, since the price of long-term debt is quite variable, what if, for example, the Fed’s future policy requires it to liquidate the debt at a loss? Won’t that have an effect? Again no, because, when the Fed liquidates the debt at a loss, the Treasury can buy back the debt and retire it at a profit. What if the Fed ends up liquidating at a profit? Yet again, no effect. If the Fed can liquidate at a profit, that means the Treasury’s borrowing costs have gone up, so, in present value terms, the Treasury has a loss to offset the Fed’s profit.
So there you have it: under present circumstances, except for possible technical and psychological effects (and the tiny effect they may have on those who are on the margin between holding T-bills and cash), the Fed’s decisions about monetizing government debt are entirely inconsequential. No doubt there will come a time in the future when such decisions will once again be consequential (as they have been during most of the past), but for all we know, that time may be a long way off.
So my advice is, ignore all the information you get about the Fed’s actions (and contemplated actions for the immediate future) with respect to the monetization of government debt. That does mean that you should ignore (or at least reinterpret) most of what I said in my earlier post on the subject. (I plan to expand on it in a future post, because I still think it has some potential substance.) Pay attention, perhaps, to what the Fed does (and it has been doing quite a lot) with private sector debt, since there we are no longer dealing with mere book-entries between the Treasury and the Fed, and real gains and losses are possible, with real effects on both public finance and private sector wealth.
But bear one thing in mind when you do pay attention to the Fed’s monetization of private sector debt – and the Treasury’s bailouts or speculative actions with respect to private sector entities. Consider the implications of the argument I have made here. The Fed’s decisions about monetizing Treasury debt make no difference. Therefore, when the Treasury does a so-called bailout, it would make no difference whether that bailout were financed by the public or by the Fed. Therefore it might as well be financed by the Fed. Therefore Treasury bailouts are no different than the Fed’s monetization of private sector assets directly. I plan, in a future post, to argue that those bailouts/monetizations are not as dangerous as some economists think (and certainly not as costly in “expected value” terms as much of “Main Street” seems to think). But bear in mind the equivalence. If you must worry about something, don’t worry about the $400 billion or so that the Treasury has used; worry about the trillions that the Fed is using.