Duration Risk: Why "Tapering" Is a Mirage and the Fed May Be Loosening Instead

John Butler made an interesting comment in his recent interview with the Financial Sense Newshour. He said, when you look at the Fed’s balance sheet, one can argue that the Fed really isn’t pulling back on its stimulus to the markets at all. In fact, Butler explained, it may actually be loosening instead:

“The Fed may be tapering its purchases in nominal dollar amounts but in terms of the amount of interest rate risk it is assuming vis-à-vis the private commercial banking system, actually the Fed is continuing to assume additional interest rate risk at an elevated rate. And so from a bank balance sheet or liquidity management perspective, Fed policy is no tighter today than it was last year. And I can’t stress this point enough: It’s just as loose today as it was pre-taper…and arguably looser depending on how you define interest rate risk.”

In case you don’t understand the above, what Butler is saying is that while the Fed is buying less and less debt each month in dollar terms, it is simultaneously buying a larger amount of long-dated bonds, which has the overall effect of keeping conditions loose.

Here’s a recent chart from the St. Louis Fed showing how they’ve ramped up their purchase of long-dated Treasuries (ten years or more) over the last few years, while pulling back on shorter maturities:

Source: St. Louis Federal Reserve

With the above in mind, Butler stated that tapering "is something of a mirage…as long as the Fed continues to absorb more and more long-dated bonds, the banks will have an incentive to increase lending and leverage, notwithstanding the ‘taper’.”

Via email, Butler explained in more detail why this is so:

Ever since the 'taper' talk began I have been making the point that by only looking at the face amounts of debt purchased by the Fed we do not get the full picture. Indeed, recall when the Fed made a big deal about long-maturity purchases in the first place? Well, they were doing so to emphasize that they still had massive firepower to stimulate credit by buying long-dated bonds. 'Duration' is a fancy word for 'maturity' which incorporates the specific coupon and principal repayment schedule of a given bond or portfolio of bonds. More subtle is the concept of 'convexity' risk, which is only significant in longer 10y+ maturity bonds as it increases non-linearly with maturity.

Consider what banks are: they are financial intermediaries that borrow short to lend long. This means they are exposed not only to the level of interest rates, but to the term structure. The Fed has the power, through debt purchases, to affect both. In short maturities, however, there is little 'duration' or 'convexity', so the nominal amount of debt the Fed purchases is a proxy for the amount of interest rate risk it assumes from the banks. The less risk, the more long-term lending the banks can make, and the greater the implied, potential 'money multiplier' they can generate.

In long maturities (beyond 10 years), the Fed takes far more 'duration' and 'convexity risk’ off the banks, yet the banks receive the new money and can lend it out nevertheless. So the risk/reward for banks can shift dramatically when the Fed purchases 10y+ bonds instead of, say, 1y bonds, and the flexibility here can easily compensate or even overcompensate for the 'taper' currently underway. And the data show the Fed is continuing to increase its purchases of 10y+ bonds, the ones that really take the bulk of interest rate risk off the banks.

So based on this, you can argue fairly that the 'taper' is something of a mirage. Bank risk and liquidity managers know this and they think in 'duration' and 'convexity' terms. As long as the Fed is continues to absorb more and more 10y+ supply, the banks will have an incentive to increase lending and leverage, notwithstanding the 'taper'.

I'm a bit surprised more has not been written about this. There was an article in the FT two years or so back making this point but now there is an odd silence. Has any US bank published on this? If not, why not? Do they not want to? Does the Fed not want them to? If the Congress or others learned that the 'taper' really isn't a 'taper' at all, would there be repercussions? I don't know, but if the 'taper' was a PR campaign [see story] rather than a true change in policy all along then this would make more sense.

In the FT article, Brian Sack, executive vice president of the Federal Reserve Bank of New York, explained how the Fed views this very process of increasing the duration risk of its portfolio as an important aspect of stimulus:

If economic developments lead the FOMC to seek additional policy accommodation, it has several policy options open to it... One option is to expand the balance sheet further through additional asset purchases, with the just-completed purchase program presenting one possible approach. Another option involves shifting the composition of the SOMA [System Open Market Account] portfolio rather than expanding its size. As noted earlier, a sizable portion of the additional risk that the SOMA portfolio has assumed to date came from a lengthening of its maturity, suggesting that the composition of the portfolio can be used as an important variable for affecting the degree of policy stimulus.

Later in the interview, Butler noted that the reason the market probably hasn't reacted that much to the "taper" so far is precisely for this reason:

"[Those] who have a direct line to the Fed, who are involved in the whole primary dealer liquidity distribution network for Treasury securities and swaps and everything else, these people are well aware that conditions have not materially tightened and their institutions can act accordingly when they receive those internal reports from people actually in the know. Say one thing, do another. That’s kind of the way the Fed is operating. And that may help to explain why the U.S. equity market has still been able to shrug off a lot of what’s been taking equity markets down elsewhere in the world."