Thursday, July 15, 2010

FOMC Minutes, June 22-23

FOMC minutes for the most recent meeting were released yesterday. Two interesting issues relate to the composition and size of the asset side of the Fed's balance sheet, and to the future path for inflation and what to do about it.

Balance Sheet Issues
Early in minutes, there is reference to a person I did not know about:
The Manager of the System Open Market Account (SOMA) reported on developments in domestic and foreign financial markets during the period since the Committee met on April 27-28, 2010.
The Manager of the SOMA holds a position at the New York Fed, and is appointed by the FOMC. He/she attends FOMC meetings, and on this occasion was called on for an extensive discussion of Fed balance sheet issues. Currently, the composition of the Fed's assets is very unusual. The average maturity of Treasuries held by the Fed is much longer than would have been normal in the past, and the Fed is holding a very large quantity of mortgage-backed securities (MBS) and agency securities, most of which are long maturity assets. Normal practice for long-maturity treasury securities held by the Fed in the past was to roll these over as they matured according to a fixed rule, as described in the minutes. The SOMA manager described some alternative scenarios for altering the fixed rule, these ideas were batted around, and the FOMC opted to stick to the fixed rule.

What's going on here? If we focus solely on Treasury securities, one aspect of the Fed's recent policy has been to move much more heavily into long-maturity Treasuries. Why? The Fed appears to believe that it can manipulate the term structure of interest rates. Short rates cannot go much lower, so it has attempted to force long rates down by buying long Treasuries. Could this be successful? Traditional central banking lore says no. Central bankers appear to operate on the belief that central bank actions have their effects in terms of returns on short-maturity assets, with little or no effect at the long end. Under the current circumstances, if the private sector can arbitrage across maturities, a swap of reserves for long Treasuries (or short Treasuries for long Treasuries) should not matter. Moving long rates down would have to depend on some inability to arbitrage, and I'm not sure that there is any evidence that such a friction exists.

Now suppose that a movement towards long Treasuries by the Fed is irrelevant. Does that mean it's innocuous? The Fed is now intermediating across maturities - the liabilities (currency and reserves) are all short maturity, and the assets go up to 30-year maturities. As is well-known, during the financial crisis, some financial institutions got into trouble due to a mismatch between the maturity of assets and liabilities, for example Gary Gorton is fond of a story about "repo runs." But the Fed is obviously not subject to runs. Currency is not redeemable in anything, and reserves can be withdrawn as currency, which is not redeemable. What is the maturity risk for the Fed? As the economy recovers, returns on alternative assets to reserves will start to look more attractive for banks. In order to control inflation, the Fed will have to raise the interest rate on reserves to make reserves more competitive relative to other assets. Now, while the Fed is currently making record profits (see here), they would not be if short-term interest rates had to rise significantly. What difference does this make? Profits made on Fed intermediation activities are just returned to the Treasury. Why would it matter if these profits are lower? Less revenue for the Treasury means they have to make it up somehow, and there is nothing that suggests that more taxation is in the wind. Thus, this means the Treasury has to issue more claims to pay US dollars in the future. If we think that what matters for long-run inflation is the growth rate in total (consolidated) government nominal debt, then this has to come back to bite the Fed in some fashion.

The maturity issue also matters in terms of the Fed's large holdings of MBS and agency securities. While there are clearly concerns on the FOMC about this, the minutes make clear that any decisions about addressing the maturity mismatch issue, or selling MBS, will be left for another day.

In the minutes, we have this:
Over the medium term, participants saw both upside and downside risks to inflation. Several participants noted that a continuation of lower-than-expected inflation and high unemployment could eventually lead to a downward movement in inflation expectations that would reinforce disinflationary pressures. By contrast, a few participants noted the possibility that a potentially unsustainable fiscal position and the size of the Federal Reserve's balance sheet could boost inflation expectations and actual inflation over time.
I'm assuming that the concerns about deflation were coming in part from Eric Rosengren. His recent interview with the WSJ sounds a lot like Krugman. In the background is a "deflationary trap" model. The notion seems to be that we can get sucked into a black hole of deflation in which we can be lost for a decade or so, just like Japan in the 1990s. This makes absolutely no sense to me, but I'll have to investigate further - I'll let you know what I find out.

Another section in the minutes gives us this:
However, members noted that in addition to continuing to develop and test instruments to exit from the period of unusually accommodative monetary policy, the Committee would need to consider whether further policy stimulus might become appropriate if the outlook were to worsen appreciably.
As I've commented before, the "instruments" in question here, the term deposit facility and reverse repos, accomplish nothing in terms of "exit." Further, as I discussed here, the Fed really has no room to move. It could buy more MBS, which if it has any effect, serves only to reallocate credit toward housing from more productive alternatives, and buying more Treasuries does essentially nothing under the current regime.


  1. The manager of SOMA is about the first person mentioned in the minutes of every Fed meeting for the past few decades or so.

    Reduce the supply of long Treasuries and you'll tend to have an effect on yield. Arbitrage doesn't mean prices don't respond to supply.

    Zero cost currency is a partial buffer for balance sheet interest rate risk.

  2. 1. Yes, apparently I wasn't paying much attention.
    2. You'll have to explain this one. If the Fed swaps short maturity Treasuries for long maturity ones, why doesn't the private sector undo any effects this would have on the term structure? Why doesn't Modigliani-Miller work here?
    3. Sure. The Fed has a monopoly on supplying currency, and thus it profits from issuing currency and buying other assets. My argument didn't involve the Fed making negative profits, just having profits go down. How much profits go down of course depends.

  3. This paper by Chris Neely:

    suggests that the Fed's purchases of Treasuries (and other stuff) with long maturity brought down long-term rates. It would be interesting to see your thoughts on this finding.

  4. “You'll have to explain this one.”

    I’m not sure what it means to suggest the private sector undo the effects on term structure or why you would suggest or expect that. The private sector can’t replace the credit and liquidity characteristics of a US treasury. And I’m afraid I don’t understand how MM is applicable in this context.

  5. Modigliani-Miller gives us a theorem that tells us under what conditions the financial structure of the firm doesn't matter - and it doesn't matter because the agents holding the firm's debt and equity essentially undo the effects of changes in the firm's debt/equity ratio. All theorems concerning the irrelevance of government policies are essentially of the same character (for example, this one: Wallace's "Modigliani-Miller Theorem for Open Market Operations" gives us conditions under which open market operations are irrelevant. This gives us a starting point for thinking about why open market operations matter. Similarly, if someone is telling me that it matters if the Fed swaps short Treasuries for long Treasuries, this must be due to the failure of some MM theorem (the financial of the structure of the Fed matters), and I want to know exactly what the failure is.

  6. I’m not an academic and not familiar with that application of MM. My intuition on MM is that a more highly leveraged debt-equity structure requires a higher cost of equity for a reduced equity component, such that the blended total cost of debt and equity capital remains the same, etc.

    The intuition isn’t as obvious to me for government debt because there is no equity account. I assume it means that different mixes of short and long term debt result in the same overall cost of debt. But the risk profile is much different. Insolvency risk isn’t comparable to the private sector case. The main risk on government debt is interest rate risk. There is no equity account, so realized risk on the cost of debt gets absorbed into future budgets and future debt rather than the realized cost of equity capital in the current period. The Fed and the government are commingled economically, so I don’t see any separate issue for the Fed’s balance sheet in this regard. Its (equity) capital account rolls up economically into the government budget and deficit.

    One other particular aspect is that term treasuries, because of their liquidity and credit quality, are quite valuable in hedging duration risk of financial institutions and convexity risk on refinancing activity in the mortgage business. So there’s supply value there that may not show up in interest rate arbitrage arguments that don’t consider this benefit directly.

  7. The value in the Fed swapping short for long maturity Treasuries is that private sector borrowing rates are typically keyed off longer-term Treasury rates and so flattening the term structure would encourage corporates/consumers to borrow and invest/consume.

  8. Robin Greenwood of the Harvard Business School has been pursuing, in what I think is a thoughtful way, the idea that relative quantities of long and short bonds matter (as do wealth flows to different agents). Here's a link to one of his papers:

  9. Ok, late to this one, but let's start here:

    "Under the current circumstances, if the private sector can arbitrage across maturities, a swap of reserves for long Treasuries (or short Treasuries for long Treasuries) should not matter. Moving long rates down would have to depend on some inability to arbitrage, and I'm not sure that there is any evidence that such a friction exists."

    Allright, my ABD is in finance, and I've spent a lot of time on arbitrage (and "arbitrage", where there's really some risk, and/or an initial your-own-capital requirement, and/or something else that doesn't fit the classical definition of arbitrage).

    Now what arbitrage, or "arbitrage", are you thinking of?

    POSSIBILITY 1– Something like this:

    The Fed tries to buy down 10 year treasury rates (say currently at 3%), by purchasing lots of 10 year treasuries. So their rate (yield) then starts to drop. So, some people start saying, hey, it's not a good deal anymore, so I'll "arbitrage" by selling 10 year treasuries, and instead engaging in a plan to buy 3-month treasuries for the next 40 three month periods.

    Here's the problem: It's not really an arbitrage because 40 three-month treasuries in a row is not exactly equivalent to one ten-year treasury. With the ten-year treasury you're guaranteed an interest rate for ten full years, not just three months. If you switch to 40 three-month treasuries you take a risk of ending up with less interest after ten years than you would have gotten with the ten-year treasury. If there's risk, then it's not an arbitrage, and people won't necessarily do it unlimitedly as long as it exists.

    If the Fed starts to push down the ten-year rate with its buying, then yes, some people will engage in the above "arbitrage", but others will say it's not worth the risk. Even if the rate on ten-year treasuries drops a half point to 2.5%, it's still worth it to me to have that ten-year lock – which no other investment, combination of investments, synthetic plan, nothing else, can provide exactly, with the U.S. Government guarantee of payment.

    Thus, if the Fed wants to push the ten-year treasury rate down to 2.5% then all it has to do is buy out every investor whose reservation rate is over 2.5%. The remaining investors (and there will be some, who care enough about getting the ten year lock and know that they can't get it exactly with any substitute) will hold at an equilibrium rate of 2.5%.

    POSSIBILITY 2 – People "arbitrage" by selling U.S. ten-year bonds and buying the ten-year bonds of large foreign governments. Again, not an arbitrage, the risks of the bonds are not exactly equivalent. But in any case, this would just slow a drop in the ten-year rate, as the Fed would be pushing down the rate in a larger market, the global one for "risk-free" ten-year fixed bonds. The Fed could still push down the rate substantially, it's just that they would have to do it for the larger global market, so it would take more buying.

    So, I don't see it as a matter of frictions. It looks to me like there just aren't any real arbitrages, just "arbitrages". Thus, the Fed can buy down long term rates with enough buying.

  10. Thanks for the link to Chari's testimony. Though it was primarily concerned with the state of modern macro theory, I think it would have been interesting for him to also examine some of the less technical policy/data analysis that some of its practitioners issued in the wake of the collapse of Lehman. For example, I wonder if he would have told the committee that he still stands by this analysis:

  11. Yes, that looks pretty funny now, doesn't it? At the time Chari seemed convinced that the people at the Board were trying to put one over on us.