In The Year 3000: Predicting The Liability Side Of The Fed's Balance Sheet
When it comes to the asset side of the Federal Reserve's balance sheet, there are no secrets: with the winddown of the bulk of the Fed's emergency liquidity programs by February 1, the majority of the Fed's current $2.2 trillion in assets will continue being outright-held securities. And even as the emergency programs sunset, the quasi-permanent, QE remnants will be here to stay. What we know for certain is that the current $1.8 trillion in Treasuries and MBS will rise to at least $2.2 trillion, as the balance of QE round 1 is exhausted. Will this purchasing of outright securities end there? Hardly. As the Fed is the only market for MBS, and as the MBS market can not allow a dramatic rise in 30 year mortgage rates, which is precisely what will happen if the buyer of first resort disappears, we fully expect some form of QE to show up and grab the baton where QE 1.0 ends. In fact just today, Fed economist Wayne Passmore, under the aegis of Atlanta Fed president Dennis Lockhart, stated during the annual American Economic Association meeting that GSE ABS should have an outright explicit guarantee by the Federal Reserve. Forget about QE then - this would be an onboarding of over $6 trillion in various assets of dubious worth, which currently exist in the limbo of semi-Fed guaranteed securities, yet which have an implicit guarantee. Of course, should the broader Fed listen to young master Passmore, look for John Williams' expectation of hyperinflation as soon as 2010 to be very promptly met. The danger of the Fed's next unpredictable step is so great that it is even causing insomnia for none other than BlackRock big man Larry Fink, who asks rhetorically "Are they going to kill the housing market?" Well Larry, unless the Wall Street lobby hustles, and the Fed isn't forced to print another cool trillion under the guise of Mutual Assured Destruction, they very well might.
So now that we (don't) know about the assets, what about that much less discussed topic: the Fed's liabilities?
As it stands now, and as we have often pointed out, the liabilities of the Federal Reserve are rather straightforward: the major items are currency in circulation (about a $800 billion, and excess reserves of roughly $1.2 trillion of overnight deposits/excess reserves. The balance is a nominal amount of reverse repos and everything else. With the velocity of money collapsing, one thing is certain: the currency (FRN) number will remain flat for a long time, unless US consumers finally start borrowing, and circulate reserves. Alas, that is not happening: as Bloomberg points out today, loans and leases of commercial banks have declined from $7.2 trillion a year ago to $6.8 trillion in November. So no, reserves aren't going down any time soon, and neither is the consumer levering up for the foreseeable future. And somehow the Fed is expected to raise rates while excess reserves are over a trillion? Hmm, empirically that probably has worked before... in Norway and New Zealand. And our monetary system and economy shares so much with both of those countries. We hope economists are already writing the textbook on this one.
Ok, fine. So excess reserves going ever higher is one option, even if they earn a minuscule 0.25% of taxpayer money (hell, the banks are doing such a bang up job just by staying alive - we sure as hell should be paying them: after all, can you imagine the eulogy that Walt Whitman (were he alive today) would write for Goldman's funeral?). Alas, 0.25% on $1.2 trillion is not nearly enough for banks to repeat the stellar year they had in 2009, so those long-term fin puts are looking ever so juicier. And yet, with all the posturing lately about reverse repos and term deposits, we are confident that very soon, probably within 2-3 months, reverse repos and TDs will start becoming an ever more dominant part of the Fed's liabilities. However, even assuming that the Fed finally manages to pull off more than a test reverse repo, for more than $200 million, with something just a "tad" riskier than Treasuries, would it really make a big difference for the money printers? We doubt it, and observations in a recent report by Nomura seem to agree with this.
The Fed is also likely to change the composition of its liabilities in 2010, but from the perspective of the broader economy we judge this is not particularly important. The Fed’s authority to pay interest on reserves allows it to influence short-term interest rates even if reserves remain well above minimum required levels. Hiking interest rates while excess reserves remain high would be unusual, but other central banks (e.g., in Norway and New Zealand) have operated under this type of system for some time. Changing the composition of the Fed’s liabilities is also unlikely to constrain credit creation. Instead of holding overnight deposits with the Fed, banks will own repurchase agreements (repos), Treasury Supplementary Financing Program (SFP) bills, and/or term deposits. Substituting one short-term, low-yielding, risk-free asset on banks’ balance sheets for another should not make banks any less willing to lend, or customers less willing to borrow. The Fed’s reserve draining operation could help control the spread between the fed funds rate and the excess reserves rate, but would accomplish little more, in our view.
And a conservative projection of the Fed's balance sheet from Nomura:
So even while the Fed struggles to convert excess reserves to reverse repos, the broader composition of the Fed's liabilities will likely not be too relevant, except to point out that currency in circulation will be flat. This should be a big warning sign to inflationists, as two thirds of the monetary base will never see US consumers, and with wages and hours worked continuing to fall, a major driver of inflationary pressures is just not there.
So back to the assets. As mentioned previously, one thing we know for sure is that the assets will likely grow at least by another $3-400 billion. What happens then is still anyone's guess (even Goldman does not have the answer, or at least hasn't decided what outcome will benefit the prop desk the most yet). So here is our take. As we pointed out a few days ago the Freddie 30 year fixed has started to rumble, and hit a 4 month high of 5.14%. Of course, actual end-consumer rates are markedly higher. Why is this dangerous? While a topic of a future post, interest rates are the biggest household net worth killer, much more so than the stock market. For instance, the lifetime payments of a $100,000 30 year fixed mortgage (with standard terms) comes down to $193k at a 5% rate... and $316k at 10%! Netting out, this implies a nearly 40% loss in the worth of the home to compensate merely for the greater interest outflows, all else equal. (the same lifetime payments for a $100k mortgage at 5% are about equal to a $60k mortgage at 10%). Of course, this analysis is simplistic and avoids discounting, but optically this is precisely what the biggest threat from higher interest rates is to the household balance sheet. And as the bulk of American net worth is not in the stock market, but merely in their home, the adverse impact from a rise in interest rates by 5, 4 or even just 3% will undo all the hard fought 2009 stock market gains (not to mention what an additional 40% loss in home values will do to bank balance sheets).
So is a 8% 30 year mortgage inconceivable? Not if you believe Morgan Stanley. And what will a roughly 3% rate increase do to house value? Not much: just take another 16% or so off current prices. Enough to finally and terminally blow up the GSEs and who knows how many banks (at this point reading bank balance sheets is utterly pointless: the FASB has made any attempt at gleaning what the true state of a bank'a capitalization is a fool's errand, which is why we love reading the opinions of permabulls who think they have some sense of just how banks are doing: they tend to be very amusing).
What does this all mean for the Fed? In our opinion, at least another $1 trillion of MBS/agency purchases, to resume shortly after QE 1.0 ends, mortgages skyrocket and Bernanke gets a rude awakening in the form of some unpleasant phone calls from certain Administration/Wall Street officials. This means that the asset side of the Fed balance sheet will likely reach $3.5 trillion by the end of 2010. And at that point, with the dollar, euro and yen in a true race for the bottom, who really cares about a few hundred billion in reverse repos between friends. Alas, we (and in this case, not the editorial version) will have much, much bigger things to worry about at that point.