The Fed's $1.25 Trillion Gambit

The Fed’s $1.25 Trillion Gambit January 14, 2010, David Kotok, Chairman & Chief Investment Officer

Abstract: This 3000-word paper offers observations about the Fed’s $1.25 trillion mortgage purchase program, the Fed’s “profits” of $46 billion, and the policy issues involved in them. It also comments on Fed interest-rate policy and the Fed’s forthcoming possible difficulties. We thank our colleagues at Cumberland, and especially Bob Eisenbeis, for assistance. We also thank Jim Bianco’s research service, Ned Davis, BCA Research, and Barclays Capital for data. We are solely responsible for any errors.

The press is trumpeting the fact that the Fed has just turned over $46 billion in “profit” to the Treasury. Much of this money originated with the expansion of the Fed’s balance sheet and the efforts the Fed has undertaken to “rescue” the American economy. By March 31, and as part of this rescue process, the Fed will complete the purchase of $1.25 trillion of federal agency mortgage paper, so the prospects are that the Fed’s “profit” will be even bigger in 2010.

The math is straightforward. Let’s assume that these mortgage assets are yielding the Fed 4.5%. That is about the market yield on GSE paper trading at durations that approximate mortgages. Is this a real profit and something to trumpet, or is the Fed actually playing with fire – and is there a dark side to this policy?

First let’s work through the mechanics of what the Fed is doing. The Fed purchases mortgage assets either from broker-dealers or Freddie and Fannie directly, paying for them with checks drawn upon itself. In effect, the Fed prints the money, since the checks are redeposited by the recipients into their bank accounts and the banks then hold those dollars as excess reserves. To the extent that these deposits are now in the hands of the private sector, the money supply as well as the monetary base (the amount of high-powered bank reserves) has increased by the amount of the purchases. The banks may invest those funds, lend them out to other borrowers, or hold them as interest-bearing deposits at the Fed. Regardless, those banks are incented to try to find a yield.

If they cannot find acceptable investments or if they are constrained by other problems, the banks need to park those excess reserves in a riskless place which does not impair their capital structure. They could purchase very short-term Treasury securities at yields next to zero or they might exercise their option of keeping those dollars back at the Federal Reserve, where currently they will be paid an interest rate of 0.25%. Regardless of what a single bank will do with the funds, the money ends up back at the Fed in some bank's reserve deposit account, because the transfer of monies between banks ultimately ends up leaving the amount of excess reserves in the banking system unchanged.

Let’s assume that the full $1.25 trillion of mortgage-backed securities is added to the Fed's balance sheet, which means that the funds used to purchase them become reserve deposits at the Fed. As long as the funds remain as excess reserves, the monetary base has increased; but the money multiplier has not kicked in, and the Fed now has mortgage assets on its books and an offsetting liability in the form of reserve deposits on which it is paying a small interest.

What do we mean by the money multiplier? A bank must hold $1 in reserves for every $10 in demand deposits. That means the banking system potentially could increase outstanding loans and deposits by $12.5 trillion if the Fed follows through on its purchase of mortgage-backed securities and the banking system chooses to make loans rather than hold reserves idle as deposits at the Fed. (Note that this is only an approximate lower bound on the multiplier, since there is now a zero reserve requirement on time deposits, and sellers of mortgage securities could choose to hold the proceeds in time deposits rather than demand deposits.) When Fed officials “worry” about future inflation risk, they are doing so because of this potential money multiplier and the risk that the amount of money chasing goods could suddenly increase. That is why they will say that although inflation seems “anchored” now, future inflation risk is high, if the present policy is continued indefinitely.

To sum this up: On April 1 the Fed will hold an asset of $1.25 trillion earning 4.5% and will be paying an interest rate of 0.25% on a liability that it has constructed to fund the asset purchase. The Fed will then make a gross profit on the 4.25% spread between what it earns and what it pays. The Fed will then subtract all its operational costs and remit the net profit to the US Treasury. Note that this process works for all Fed assets and explains how the Fed has “earned” $46.1 billion in the last year. But before one thinks that the Fed has suddenly become a profit center for the Treasury, one must also consider the risks associated with that strategy.

By becoming the buyer of last resort of mortgage paper, the Fed has supported the housing-market stabilization policy of the Obama administration and done so by printing money. The Fed’s buying raised the price of those securities, which had the effect of lowering their yields from what they would otherwise be. Various analysts suggest that the approximate 5% Fannie conforming home mortgage rate would have been well above 6% if the Fed had done nothing. One Fed official (Boston Fed President Rosengren) suggested that the home mortgage rate will rise 0.75% when the Fed ends the program in March.

Additionally, Fannie, Freddie, and FHA are now the funding sources for nearly all home mortgages in the United States, and the private-sector home-mortgage lending business has been nearly eliminated by the use of federal credit applied through the federal agencies. Also note that these same agencies have managed to lose over $100 billion so far and are now finally booking more of those losses. Jim Bianco reported this item today: "Laurie Goodman of Amherst Securities (formerly of UBS) is a well-known expert on securitization and MBS. In a 17-page report out yesterday, she estimates the losses at Fannie/Freddie at $448 billion.” And also note that Congress is responsible for creating this federal-agency mortgage mess and that the blame for the program is best directed at the Congressional leadership.

The Fed had cleaner hands here before it launched into this massive purchase of mortgage paper. Criticism of the Fed centered on its low interest-rate policy under Greenspan as a way to allow housing speculation. There is some evidence that the Fed did contribute to the housing bubble when very low teaser rates on adjustable-rate mortgages encouraged speculation by home buyers who believed housing prices would only rise. Chairman Bernanke’s defense of the Fed in his January 3 speech was not received as a convincing counterargument. And his call for regulation as the answer to bubble avoidance was also greeted with skepticism. He ended with this:

“My objective today has been to review the evidence on the link between monetary policy in the early part of the last decade and the rapid rise in house prices that occurred at roughly the same time. The direct linkages, at least, are weak.”

We wonder.

If the linkages are weak there should be little impact from the Fed’s $1.25 trillion purchase of residential mortgage paper. If weak, why are they doing it? If the linkages are strong, why were they weak in the past and strong now? Regrettably, Chairman Bernanke’s extensive and serious discussion of technical policy rules, like John Taylor’s eponymous, one only added to the confusion about Fed policy. Chairman Bernanke’s speech didn’t clear up the issues about the Fed and the financing of housing lending.

The largest part of the Fed’s current balance sheet is committed to holdings of GSE-based residential mortgages. The Fed has now become complicit to the extent that it has subjugated its balance sheet to the Treasury and Congress in the support of housing. That is where the Fed has added the risk of an explosive expansion of the money supply should banks decide to lend. That risk is not imminent but it is developing. We believe that when it manifests itself, it will do so in a profound and volatile way.

There is much debate about how and when the Fed will “unwind” or “exit” this program. Some argue that the Fed will be selling some of this $1.25 trillion. We think that may occur, but only many years from now and then only in small increments and after the market has stabilized. Fed Vice-Chairman Don Kohn has alluded to the fact that the Fed must unwind slowly and will quickly back off if markets show any pressure.

The Fed is loathe to admit errors in policy. And they do not like it when they are criticized for failing to be transparent about policy. As outsiders we do not have this constraint.

In our opinion, we think the Fed is now trapped.

By becoming the buyer of last resort, the Fed has now impacted the markets in such a way that the very idea that it may withdraw has caused mortgage interest rates to rise. Markets aren't dumb, and they realize that rates will rise, for two reasons. First, if the supply of funds to Freddie and Fannie stops with the Fed's purchases, then home-mortgage interest rates will have to rise. Moreover, they will rise even further if the Fed starts selling its existing securities into the market. What this also means is that the interest-rate risk associated with any future increases in interest rates will be shifted from the private sector to the Fed and ultimately the taxpayer – and this risk will grow as the Fed begins to unwind its current low-interest-rate policy.

Note that mortgage rates have already risen recently, and the rise occurred while the Fed is still in buying mode. If the Fed plans on doing any selling it will have to reveal those plans, and markets will immediately change their pricing to reflect the possibility of this huge seller entering and dumping mortgages on the market. That expectation alone will cause prices to fall and rates to increase in a way that could be very serious. In other words, the Fed has trapped itself, since it will incur capital losses if it sells mortgages into a rising interest-rate environment, which would only put further upward pressure on rates and risk damaging the mortgage market.

This dilemma suggests that the Fed is going to be a long-term holder of this mortgage paper. It might serve the market and give the Fed more credibility if the Fed were to make that clear to the market. Such clarity would remove any “exit risk” pricing premium currently imbedded in mortgage interest rates.

So if the Fed will not be a seller of these mortgages, another option, and what now seems to be emerging as a preferred one, is to avoid future inflation risk by raising the interest rate on reserve deposits. The Fed has already constructed these deposits to fund the purchase of this mortgage paper. These deposits present a riskless option to banks to keep funds on deposit with the Fed rather than lend them out for a higher, but riskier return. Maneuvering the rate paid to banks on these deposits may be a way to neutralize the monetary impact when the economy eventually begins to recover in a sustainable way.

The Fed can raise this reserve-deposit rate at any time and can set it anywhere it wants. It will likely change the rate only when the Fed decides that it wants to keep banks from buying other assets or expanding loans too quickly. The Fed has already said that this is part of its new package of monetary policy tools. Furthermore, they are going to expand this tool by engaging in term-deposit auctions in addition to the existing overnight reserve-deposit structure. This will change the way they implement the new structure of the liability side of their balance sheet. Term-deposit auctions may become the critical item for Fed watchers to examine. The are likely to commence in March.

Of course, manipulating the rate on excess reserves doesn’t withdraw liquidity from the banking system, as asset sales would; it only temporarily limits the effective size of the money multiplier. The longer the term-deposit auctions and the greater their size, the more the Fed can extend the “temporary” character of this program. We believe they may have to do it for a very long time.

Raising or lowering the reserve-deposit rate does not directly affect the existing $1.25 trillion of mortgage assets held at the Fed. It does affect the extent to which banks seek higher rates by making loans, and it does increase or decrease the money supply as banks decide whether or not to hold funds idle in the Fed. The direct impact on mortgages occurred when the Fed made the initial buys.

The reserve-deposit interest rate will now function just like the raising and lowering of any other policy interest rate. Raise it and some types of potential lending are constrained; lower it and something is encouraged. Since the Fed starts at near zero level, the trend must necessarily be up in the beginning.

What is the cost-benefit of this program? Remember, the Fed is going to “make” about $50 billion in the first twelve months following the full implementation of the program. And remember too, it will give that $50 billion to the US Treasury. So, in effect, the Treasury (taxpayer) got a $50 billion subsidy from the Fed for doing nothing, while the beneficiaries of the mortgage-financing apparatus got a lower interest rate on each of their mortgages.

These benefit recipients included some relatively new home purchasers. The main beneficiaries were those who re-financed an existing mortgage, along with those homeowners who were trying to sell houses in a market where prices were falling, and the lower mortgage rates allowed marginal buyers to obtain financing of their purchases.

So far all we see is the benefits of this program. Can we estimate the cost, since we know that there has to be one? If so, was the estimated cost worth it and are we getting the result that policy seeks to accomplish?

Is a $50 billion annual subsidy of this selected group of home-mortgage borrowers enough to stabilize the housing market? That asset class has declined in value from over $23 trillion to under $18 trillion.

Remember: this $50 billion is an annual subsidy that went directly to the benefit of about 5 million new and existing home-sale transactions, about 80% of which were financed with federal agency mortgages. We could make some assumptions about repeating this $50 billion for each and every year those mortgage users will be around. We could derive a present value of the subsidy. But the duration of the subsidy is uncertain due to the lack of clarity by policy makers.

One method of cost estimation might be to determine how many mortgages were newly placed or refinanced during those few months when the Fed induced a lower interest rate. We cannot know that yet, since the program is still ongoing, so let’s just venture some assumptions and let the academics finalize the cost estimate a year from now, after the program ends and the numbers are available.

We will assume that the cost of temporarily lowering this home-mortgage interest rate is equal to the benefit. Thus we can use the $50 billion number as the ANNUAL cost. Since that cost will be deferred until the Fed raises the reserve-deposit rate, we have no way of knowing what the total program cost will be over the life of the program. In theory, the Fed could hold the entire $1.25 trillion in mortgages until they all pay in full. In theory, the Fed could hold the reserve-deposit rate at 0.25% for the entire life of the program. Anyway, let’s use the $50 billion annual cost for the first year.

Let’s say there will be 4 million mortgages directly impacted by this program. We are purposefully choosing a number that is close to the expected level of activity so it avoids the criticism of underestimating. Let’s say that the average size of the house transaction is $250,000, which is probably close to accurate if we look at the distribution of conforming mortgages. It is our assumption for the average national selling price of houses during this period. At 80% loan-to-value, the average mortgage would be $200,000.

These assumptions mean the $50 billion of annual subsidy is to be divided by the 4 million direct beneficiaries. This math implies that the subsidy was $12,500 per mortgage for the first year. Somehow that strikes us as a very inefficient and costly way to try to stabilize the housing market. Of course, the numbers might be and probably are something else. But the idea seems clear enough.

Inject the policy issues into the discussion and this quagmire gets deeper. This money is really involving the Fed directly in fiscal policy. The Fed's mortgage purchases are just a back-door way of supporting particular borrowers at a particular point in time to benefit particular segments of the housing industry. Is that the intended role of the central bank?

The other consequence has been to shift the interest-rate risk of the policy to the Fed, and hence the taxpayer. The Fed will surely have to begin raising rates at some time in the future, even if it’s a long time from now. And those rates will have to increase more than proportionally as the economy improves, lender risk aversion declines, and the opportunity cost of holding excess reserves at low reserve-deposit rates increases.

Thus one can ask fairly if it is the Congressional intention for this subsidy to be directed in this way. Did the taxpayers have anything to say about it?

And lastly, and most importantly, is this the responsibility and purpose of the central bank?

We wonder again.

At Cumberland, my colleagues and I have been long-time and consistent advocates of the central bank’s independence, so that it can formulate monetary policy. We believe that the central bank’s proper role is to maintain the store-of-value characteristic of money, by keeping inflation restrained and by keeping deflation from doing its devastating damage. And the central bank must see that the money is used as the medium of exchange. Also, it must facilitate the use of money as a unit of account. These are the three things that central banks are supposed to do: store of value, medium of exchange, unit of account.

Somehow, institutionalizing a policy of subsidizing a $200,000 home mortgage with an annual stipend of $12,500 is not on that list.

We are currently scheduled to appear on the CNBC Wall St. Journal Report this Sunday night at 7:30 Eastern time, and hope to discuss some of this issue then. Local stations may carry that show at other times.

Cumberland AdvisorsSM is registered with the SEC under the Investment Advisors Act of 1940. All information contained herein is for informational purposes only and does not constitute a solicitation or offer to sell securities or investment advisory services. Such an offer can only be made in states and/or international jurisdictions where Cumberland Advisors is either registered or is a Notice Filer or where an exemption from such registration or filing is available. New accounts will not be accepted unless and until all local regulations have been satisfied. This presentation does not purport to be a complete description of our performance or investment services.

Please feel free to forward our commentaries (with proper attribution) to others who may be interested.

For a list of all equity recommendations for the past year, please contact Therese Pantalione at 856-692-6690,ext. 315. It is not our intention to state or imply in any manner that past results and profitability is an indication of future performance. All material presented is compiled from sources believed to be reliable. However, accuracy cannot be guaranteed.