Sic Semper Tyrannis

While pouring over Treasury International Capital (TIC) and Federal Funds Rate data recently, I found something that seems quite puzzling when one looks at foreign purchases of US corporate and Treasury securities, juxtaposted with the Federal Funds rate. The chart of these series seems to exhibit a striking change in relationship around 1995--most notably, the response of official purchases of Treasury securities to the Federal Funds Rate seems to have inverted:

Follow up:

(click for larger version. Note--labels on Funds rate are not negatives.)

This apparent inversion of market behavior is quite distinct, and in my humble opinion, demands explanation. One might ask whether Fed policy is what changed, or whether the market response to Fed policy (in terms of the Funds Rate) changed.

If it is the former, then in a sense one is arguing that the Fed funds rate is "reactive" and attempts to keep bond buying at a stable level (quelling excess bond buying, and stimulating low bond buying). Then the "1995" policy change would be that the Fed decided to reverse this, inflating high levels of bond buying even further, and crashing them when they were already low.

I consider this unlikely, because I don't think the Fed would intentionally try to create bubbles and boom-and-bust cycles (particularly for government income, as the case would be here). Further, I doubt the Fed is reactive towards bond purchase levels, using it as any sort of benchmark.

I think it is more likely that the market itself is reacting differently towards the Fed (with the Fed Funds Rate the independent variable), though not because of any inherent change in the market itself--rather, because of a seemingly-unrelated change in how the Fed operates.

That change is "transparency".

The transparency policy was introduced in the 1994-1996 timeframe, precisely when the bond market seems to have undergone this behavior inversion. Under the transparency policy, the Fed would issue statements with its regular board of governors meetings, summarizing the Fed's current thinking about the trajectory of the economy, as well as its plans for moves in terms of interest rates (and potentially other controls available to it, though I don't think the Fed has ever widely acknowledged any of them, most notably the reserve ratio).

This policy has been the modus operandi of the Fed ever since. While it has been hailed as a veritable "capitalist glasnost," I think there is a distinct dark side to the policy, as evidenced by the market's inversion. I think the inversion can be traced to a fundamental behavioral change in bond market participants in the presence of "transparency" of the type the Fed practices.

Imagine you are a bond buyer, pre-Fed-transparency. You have little inkling of what the Fed is going to do regarding interest rates at any given time; certainly one can make educated guesses based on macroeconomic indicators, but there never seems to be agreement on the implications of those, even amongst "experts." So for all intents and purposes, the Fed's proximate actions regarding interest rates are an unknown; and even those who think they do know will face just as many market participants who think they know the opposite, thus causing their relative influence to wash out.

So the overall behavior of bond buying upon a rate change will be very simple and intuitive: when the Fed raises interest rates, bond buying will increase, because bonds have become a more attractive asset (higher return, or yield). When the Fed lowers interest rates, bond buying will decrease, because bonds will have an inferior return. Simple.

It is important to note that in this world, Fed interest rate changes are "in the void"; they don't conclusively change one's information about the trajectory of the Fed's next move. Future moves are, in fact, almost a non-factor. Everyone may have their guesses, but no "signal" would emerge from the noise of the market.

Well, that would have all changed with the dawning of the transparency policy. Suddenly, the Fed is not only telling you what the current interest rate is, but it is telling you what the next interest rate will probably be. In fact, we can go one better: the Fed is telling you the trajectory of interest rates over the whole "near future" timespan.

Now what is a market participant going to do?

Well, when interest rates rise in an upward trajectory, people will sell bonds, overall. Why? Because in the near future, they know yields will be higher, making the bonds bought at the present time less valuable. It would be downright stupid to buy bonds during a rate increase cycle, unless the intent is to hold the bond until its expiration for a "risk-free", (relatively) low-yield return. But that kind of behavior does not predominate in bond market agents today.

When interest rates are in a down-trend, the opposite now occurs: since the yield on near-future bonds will be lower than the present yield, the value of present bonds will likely go up soon, leading to a tidy profit.

This behavior is the exact opposite of the behavior that would occur pre-transparency, and the critical element was simply to add information about the likely trajectory of the Fed interest rate series. Information is a powerful thing--particularly in markets.

I think there are also some deep ramifications of this bond-market transition for markets in general. By definition, inversion of bond-buying behavior implies that when interest rates are going up--a time normally strong for bonds--bonds would be predominantly getting sold off, as discussed. But this in turn suggests that during these times capital would in fact drain out of the bond market instead of in, and find its way into other financial instruments--most notably equities.

But if the goal in raising rates had been to slow the economy, this would obviously mark a failure in the sense that raising interest rates would now engender a stock market rally.

[And what do you know, that's just what we've seen over the past three years.]

On the flip side, lowering interest rates, which is supposed to stimulate the economy, would suck people disproportionately into the bond market, as they expect to profit purely from a trajectory of falling rates. This may also cause some diversion of funds into corporate debt, which could still act as some stimulus--think stock repurchases funded by borrowing.

The upshot is that raising rates would massively stimulate equities, and falling rates would perhaps still simulate them a lesser amount (through a more favorable corporate debnature environment). This would then amount to asymmetric stimulus emanating from the Fed, of the kind that always tends to inflate the private economy.

Consider, by the way, that "absolute" low interest rates are always attractive to debtors, so below some real lending cost threshold, private (and public?) debtors will take on essentially as much debt as they can get. If the inflation rate were to be lowballed, the situation would be even better for them. So I think there is not only some credence to arguments that the credit market behaves differently in the presence of "absolutely" low interest rates (as we have in today's era), but that this argument should be even stronger in times of very low real interest rates (which I think we also have today).

So it would appear the private securities market will always manage to get inflated in this regime. Indeed, in the above chart, one can see that corporate securities purchases seem little effected by interest rates post-1995; they always seem to go up (other than responding slightly to massive official moves in Treasury securities).

Adding to this point, the following chart showing net US securities purchases of all types by foreigners, could also be a consequence of this effect:

(click for larger version.)

Note the skyward launch immediately prior to 1995. Coincidence? I'm not so sure. There has been no regime in our history in which such an extreme event occurred, and we have glaring changes in liquidity policy and now, apparently, changes in the bond market's behavior right at that point.

Thus, I suspect that the transparency policy--which seems like a good, honest, innocent idea--actually has disastrous asset-inflationary effects when applied to the kind of liquidity-based central management the Fed performs. The policy may need to be reconsidered.

Or maybe the Fed should be reconsidered.

One other point about the new behavior. Above I discussed what happens during rising or falling Funds Rate trajectories. What of intervals of a flat Funds Rate? If I am right about the fundamentals, then a flat Funds Rate interval would be biased towards the stock market, as opportunities to time-arbitrage US Treasury Securities would vanish, and speculative money would likely pour into the stock market in pursuit of higher returns. Due to the herd buying effect, these returns would likely come very quickly, providing "instant gratification" for many. Simultaneously, bonds would initially "crash," but then likely rebound a bit and stabilize, as "neutral" bond buying behavior takes over. The data appears to support this thesis (as does the recent unfolding of events).

Also, allow me to attempt to head off potential rejoinders that the bond market (and to some extent the other markets) will be controlled predominantly by inflation/recession fundamentals. One of the great victories of the Fed of the past decade seems to be having convinced everyone that, as far as financial decisions go, there is no inflation (and there won't be in the future). Everyone is well aware of the fact that general prices are still going up (and even accelerating), but somehow this notion has been severed from the financial notion of inflation. Its like people live two lives--a real economy life and a financial economy life, with the latter floating off in la-la land. Further, you have a large fraction of bond (and other) market particpants acting as intermediary investors, which means they don't have to live off of the real level of investment returns. Thus I am not sure it is very accurate to model the behavior of participants in the market--especially the bond market--as containing an "inflation expectations" component.

Finally, everyone seems to have become "recession blind", with the Fed claiming that landings will always be "soft" (recession-free) and the entire cadre of "expert" punditry never seeing a recession until after it arrives (i.e. most economists fail to see recessions coming).

In the midst of all this confusion and misinformation about recession, inflation, and their influence on the financial markets, it would seem to me that any clear signal the Fed gives about near-term interest rate behavior would overwhelmingly dominate the behavior of bond market participants.

The above analysis only a beggining; there are certainly more glaring questions. Why do these effects seem to hold even though foreign official (central bank) purchasing shouldn't be convertible between public and private securities? Does the effect hold for domestic purchasers of bonds and stocks? How far back does the pattern hold? What is the influence of inflation (both CPI and a "real" inflation rate based on M1)? Perhaps some of you can help answer these questions.