Monday, September 27, 2010

Are Reserves Running Off?

I'm not sure if there is anything important going on here. It could just be a blip. However, as you can see in the chart, reserves fell below $1 trillion for the first time since January, and there is a clear downward trend (with a downward blip). The other component of outside money, currency (in the next chart), continues to grow at a steady pace of about 5% per year. If inflation were to increase, this is where it has to come from - reserves are shed by banks as they become relatively less attractive, and currency keeps growing. We need to see more data before attaching more significance to this, but it bears watching.


  1. "Reserves are shed by banks as they become relatively less attractive."

    Can banks "shed" reserves? Under ZIRP, there is no reason for the Fed to conduct OMO to soak up reserves. Therefore, it seems a reserve drop would have to be caused by the Fed's failure to replace maturing MBS with (settled) Treasury purchases in a timely fashion. I suppose banks could lose reserves as demand for currency rises, but I don't think that is what is going on here--deposits and required Reserves are actually rising.

    The last point above brings up an interesting development. Banks are buying Treasuries, and their Excess Reserves are falling as deposits and Required Reserves rise slightly. Are banks front-running Fed Treasury bond purchases? If so, this front-running amounts to a back-door monetization of part of the deficit, to the tune of about $26b in August and about $16b (run rate) so far in September.

    The question I ask is, if the banks decided to buy significantly more Treasury's, how would the Fed know when to raise the IOR? In other words, with low effective reserve requirements (3%?) given sweeps, it takes a whole lot of Treasury purchases (and their resulting deposit creation) to make a visible dent in ER's. By the time the Fed realizes what is going on, inflation from deficit monetization might well be on its way. Perhaps this helps explain why the CRB commodities index is at record highs...

  2. "Can banks "shed" reserves?"

    Sure. Outside money is currency plus reserves. Reserves can turn into currency one-for-one. In this case though, there has to be something else going on in the Fed's balance sheet to account for what is going on with reserves. If we take the Fed at its word, it is replacing MBS with Treasuries as the MBS run off, so it can't be that.

    "Are banks front-running Fed Treasury bond purchases?"

    I'm not quite sure what you are getting at here. If the banks buy Treasuries, I don't think of that as "monetization," though it has the effect I think you are getting at. The banks shift from reserves to Treasuries, reducing the demand for outside money, which is inflationary. For me, monetization is something the Fed does.

    Another point, which I think you are getting at is that we might as well think of reserve requirements as essentially non-binding in the United States, given the vehicles banks have for getting around them.

  3. Banks don't "shed" reserves.

    Non-bank customers of banks will increase their holdings of currency over time, which will displace an equivalent amount of excess reserves.

    But that's a trend which is not bank initiated or controlled.

  4. No. The total quantity of outside money is determined by the Fed. The composition of that stock (currency vs. reserves) is determined by the collective decisions of consumers, firms, and financial institutions. Suppose, for example that, for whatever reason, banks are not content to hold the existing stock of reserves, given interest rates and prices. Then, something has to give. Under the current regime where the interest rate on reserves determines the fed funds rate and all other short rates, it seems that what has to give ultimately is the price level. The price level rises so that banks are content with the quantity of reserves they are holding, and consumers and firms are content withe the quantity of currency they are holding. What happened? The stock of currency rose in nominal terms and the stock of reserves fell in nominal terms. Effectively, reserves where "shed" by the banking system due to something that initiated in the banking system.

  5. How the price level would rise in your example seems a bit obscure. It is hard to see how it would be initiated or controlled by banks.

    So banks don't "shed" reserves, right?

  6. Total demand for outside money in real terms falls; supply of outside money in nominal terms is fixed; price level has to rise. I hope I explained satisfactorily what I meant by "shed." They do shed.

  7. Here's something interesting. On the Fed's balance sheet, from the week ended September 15 to the week ended September 22, average assets were essentially unchanged. On the liability side, reserves went down by about $52 billion, but deposits held by the Treasury went up by about $55 billion.

  8. The increased in currency in circulation mostly came as a result of an increase in foreign demand. Do you really think this increasing demand for the US currency abroad will create inflation at home?

  9. That certainly has to be part of what is going on. That huge increase in the currency stock in late 2008 obviously wasn't producing much inflation, and likely this has something to do with some or all of the increase being held outside the United States. Of course, whatever caused that increase in overseas demand for US currency (somehow financial-crisis-related) would presumably reverse itself. Further, you say

    "The increased in currency in circulation mostly came as a result of an increase in foreign demand."

    This is not a fact. We have no idea where the stock of US currency resides.

  10. (From the first Anonymous above)

    Here is Kocherlakota from today:

    With that said, I have indicated in earlier speeches that $1 trillion of excess reserves does create a potential for high inflation at some point in the future if the FOMC does not react sufficiently fast when it starts to see inflationary pressures.

    Again, this seems to imply that the presence of $1tr in ER's is somehow more inflationary than if there were zero ER's. This seems incorrect. If banks face higher credit demand in the future, they can obtain as many "licenses" from the Fed to make loans (Kacherlakota's useful terminology) as they want. This is because, at any given FF target rate, the Fed will supply any reserves through OMO necessary to maintain the FF rate at target.

    ER's are like having solar power at your house, but its price is still controlled by the power company. Yes, the all the power you need is right there, but it is no different than drawing on the grid from the power plant at the same price for any needs. What matters in that analogy is the price of the power, not where it is located.

    So why is Kacherlakota perpetuating the idea that ER's are potentially "inflationary" compared to not having them? This confuses the markets, IMO.

    "The effects of the financial crisis on currency in circulation domestically were less significant than on currency in circulation internationally. Although domestic demand appears to have increased briefly in the fall of 2008, the increase of the FDIC deposit insurance limit and other government actions to address the crisis allayed domestic concerns and demand seems to have returned quickly to normal levels. The majority of the growth rate increase was driven by demand for $100 notes internationally. The Federal Reserve requested that the BEP accelerate the Board's print order for $100 notes during this period so that the Reserve Banks could continue meeting international demand. Although payments of $100 notes abroad have returned to normal levels, most of the currency the Reserve Banks paid into circulation during the financial crisis remains in circulation as Reserve Bank receipts from circulation remain historically weak."

    Beside, do you think that your inflation story resulting from banks shedding reserves is a fact? Any bankers reading this would laugh like hell.

  12. The Fed obviously tries to estimate how much currency is here and how much is there. The Secret Service also apparently has ways of obtaining estimates of what the currency is up to. However, by its nature, we can only make rough guess about where the currency is. You don't know whether the briefcases full of $100 bills are in Miami or Moscow.

    I'm trying to help you get this, apparently without success. Whether some bankers have a good yuck about it doesn't concern me at all.

  13. Stephen,

    From the first Anonymous:

    There seem to be a number of anonymous posters, which might create some confusion. I have been asking about the inflationary impact of Excess Reserves. If you have comments on the Kocherlakota speech, I would be interested in reading them. Thanks.

    Henceforth, Anon1

  14. anon1:

    Yes, Kocherlakota's speech is interesting. I'll write a post on it when I have the time.