Another Unnecessarily Embarrassing Day for the Fed
US Government bond securities throughout the 1930’s would trade with negative yields the closer to maturity they became. It was not then, and is still not in the modern day, uncommon for treasury bill yields to be slightly negative upon expiration, but that was and is more or less a few basis points. These were longer-dated bonds, though, and the negative yield was at times on the order of several percent.
Stephen Cecchetti in a December 1987 paper, The Case of the Negative Nominal Interest Rates, written for the NBER introduces it by reciting a New York Times bond quote from December 31, 1932. A 3 ½% Liberty Bond was listed at a yield of -1.74% on that day. Since any bond holder has the option of cash, there is outwardly no reason for this to be even possible. But, as Cecchetti recalled, “from mid-1932 through mid-1942, the vast majority of coupon bearing U.S. Government securities bore negative nominal yields as they neared maturity.”
There was obviously something of value in those bonds, otherwise they would be sold long before reaching zero yield. That demand also could not have been strictly related to the bond’s superficial characteristics. The answer was simply the primitive method by which the US Treasury Department sold its debt to the public.
There was no highly developed primary dealer system in those days. Auction processes were far more dispersed and, you might say, democratic. Treasury sales were open to pretty much anyone, provided you could put up a deposit for a substantial amount in anticipation of a debt sale whose terms you wouldn’t quite know in advance. In fact, it was usually the case that there was but two weeks’ notice before the specifications had been drawn and released.
To correct for the messiness of the process, the Treasury would essentially pay people, or financial firms, to buy the securities. It was required by law to offer debt at par, but the department would set the coupon of any new issue at a higher rate than what was implied by market prices. Therefore, it would sell its bonds at par but giving the buyer a better interest rate than par.
Absent the primary dealer system, one of the means for encouraging bids was to give preferential treatment to holders of expiring issues. Therefore, if you were the owner of a federal bond security that would give you privilege to participate in the undervalued bonds at auction replacing it as it matured. Clearly, as Cecchetti described, that was a valuable option on at times substantial free money. He called it the “exchange privilege.”
The process was not without risk on the part of the buying participant. Short notice as well as the possibility for changing market conditions left it far from sure arbitrage. It was an opportunity, one that held public interest and to which the public was left to pay.
From the government’s perspective, there was another consideration. From 1929 until 1939, the amount of outstanding debt rose substantially, from $16 billion to $41 billion as a result of intended fiscal “stimulus.” This was unique in American experience, at least outside of wartime. The Treasury Department was constantly on alert for being left to twist should interest rates get out of hand.
As the economy seemed to recover from the depths in 1932 and early 1933, that is precisely what concerned officials. In the spring of 1935, Treasury representatives held several conversations with their counterparts at the Federal Reserve aimed at convincing the latter group to buy UST securities. It presented its case as proposed policy to stabilize the government bond market for whatever risks. In the conditions of 1935, that meant what they were increasingly sure was latent inflation pressures building up underneath.
It wasn’t philosophically a hard sell to the men at the Fed. They agreed with that pronouncement, and had even judged it their main objective. It had already been discussed at that early date where the FOMC might preemptively act before inflation got out of hand. The amount of gold then flowing to the US as well as the assumed recovery for the Fed as well as the Treasury was in their combined view a threat to economic and financial viability. Banks had experienced a massive growth in bank reserves, including excess reserves that the Fed in particular considered to be nothing but a surplus. As such, they were in the central bankers’ collective estimation awaiting only a trigger to unleash runaway monetary chaos (Weimar Germany wasn’t at that time totally obscured by 1929).
Acquiescing to Treasury’s suggestion, the Fed in early 1935 bought long-term bonds for one of the first times in its history. It is difficult to assess the efficacy of the program because by and large US Treasury rates for the whole curve were falling throughout. The 10-year yield which had been 3.39% in September 1929 was 2.74% by May 1935; and 2.36% by January 1939.
Not only that, the yield curve never behaved as it was supposed to. Throughout the year 1929 it had been inverted. The difference between the 10-year note and the 3-month bill was about -100 bps to start in January, and still about -90 bps by the time the crash rolled up that October. Then, as is almost always the case during the outset of contraction, the yield curve steepened sharply as liquidity preferences triggered a run to shorter dated treasuries (leading to often rapidly declining short-term rates against more stoic though falling longer-term yields).
By the middle of 1932, bill rates had sunk all the way to zero. The 2-year note that had been trading around 4.5% in 1929, by the second half of 1932 was nearer 2.0%. That was the curve’s steepest point, right at around the bottom. By the time Treasury was querying the Fed about agreements over bond dangers, the yield curve had further flattened. That wasn’t totally out of line, as the yield curve in any recovery should flatten as it takes hold.
But it was flattening in the wrong direction. Normalcy, especially after the short end’s experience at zero and near-zero yields, is for the unusually steep curve to give way to flattening as nominal rates rise. In reverse of the steepening, short rates move up quicker than the again more tolerant long end so that the whole curve shifts upward to diminishing degrees by maturity.
In the 1930’s, it was always the opposite. Yields on bills stayed at zero aside from a few brief outbursts, while longer rates marched relentlessly lower a little bit at a time. For the decade as a whole, the US Treasury curve flattened further and further at lower and lower nominal points. Recovery projections didn’t bear out at all. Inflation was never a problem, nor was there any related difficulty for the Treasury Department ginning up demand for its paper despite continuous, large war-like deficits.
It wasn’t just UST yields that sank incrementally, as corporate credit followed closely, too. Corporate AAA-rated bonds that yielded in the aggregate around 4.7% in 1929 and 3.7% in May 1935, crossed under 3% less than four years later at the decade’s end. Even riskier Baa-rated credits were lower for the period, though not in any way taking a straight line. Yields that had averaged about 6% in May 1935 were as low as 4.5% at the outset of renewed contraction in 1937, spiking back to 6% at its end, and then falling below 5% again by the spring of 1939.
It would be many years later before a consensus could be reached on how that all could have happened. Milton Friedman put it best in 1967:
“As an empirical matter, low interest rates are a sign that monetary policy has been tight in the sense that the quantity of money has grown slowly; high interest rates are a sign that monetary policy has been easy in the sense that the quantity of money has grown rapidly. The broadest facts of experience run in precisely the opposite direction from that which the financial community and academic economists have all generally taken for granted.”
Interest rates especially on securities like UST’s are an expression of among other things liquidity preferences. These apply to individual investors as well as more so in our “modern” age to financial institutions. If money is tight as Friedman proposed, then demand for liquid instruments will become relatively greater. That would be the case even when government deficits and outstanding debt might otherwise create inordinate problems for bond markets, government budgets, and the economy through unstable and rising nominal rates.
It is therefore a simple expression of market-based risk perceptions. Which is greater, the risk of catastrophic default (or even just partial jubilee) on federal bonds or the economic consequences of further depression? In the 1930’s as well as Japan after 1990, the answer given was clearly the latter no matter how many times along the way officials were absolutely certain recovery (and inflation) was at hand.
In that way, however, the more certain officials are for no actual reason will only further counterbalance the risks toward the latter case. That’s just plain human nature. People might be enthralled by credentials for a period, but unless they are backed up by tangible proof of competence they won’t be left for much more than media editorial standards. They may be taken seriously on TV, or in newspaper pages, these authorities with official positions and pedigree, but markets are under no such obligation to treat them that way.
In August 2014, Federal Reserve Vice Chairman Stanley Fischer admitted to an audience in Sweden what was widely acknowledged but had never, ever been stated.
“Year after year we have had to explain from mid-year on why the global growth rate has been lower than predicted as little as two quarters back. Indeed, research done by my colleagues at the Federal Reserve comparing previous cases of severe recessions suggests that, even conditional on the depth and duration of the Great Recession and its association with a banking and financial crisis, the recoveries in the advanced economies have been well below average.”
Central banks in the 21st century are in their own estimation as much about psychology and credibility as money (my view is that they are all the former and none of the latter, the consequence of decades of leaning further and further in that imbalance). For the Fed’s #2 to publicly state even the obvious (and even if it was all the way over in Scandinavia) was a serious rebuke. But it was never taken that way by convention.
The timing of it could not have been more substantial. The summer of 2014 found the confluence of a number of factors that was very reminiscent of those in the mid-1930’s. Though the economy had consistently underperformed (globally), policymakers had become certain that inflation was near at hand and that risks were shifting further in that direction. As late as May 2015, Janet Yellen claimed in a speech in Providence, RI, that delaying rate hikes could risk “overheating the economy.”
The FOMC voted for one such “rate hike” in December 2015, then waited a full year for the second. In the interim, the economy fell almost into deflation (the CPI was negative in several months) and experienced a quite severe downturn. Most embarrassing for the Fed had to have been that on the very day they did finally approve that first step they also released figures on Industrial Production with a negative sign (a rare occurrence for a statistic going back to 1919 associated in every prior case with recession). Subsequent benchmark revisions now show that the downturn was for US industry deeper and started earlier – in April 2015.
For some reason, Milton Friedman’s admonishment remains unheeded. As such, people, mostly in the media, which depends still upon economists, think the bond market is some great mystery. It is common to hear now, just as in 2014, that “interest rates have nowhere to go but up.” The reasons for that statement are spurious, dependent over time on less and less. What is left for them today is solely the unemployment rate.
At 4.3% for the latest figure, that is a level that should be creating all sorts of labor distortions – of the good kind. It suggests long ago the end of “slack” so that now the economy has to be well into a labor shortage. In simple economic terms, a shortage of anything including workers leads to an often sharp increase in its price; in this case wages. As wages accelerate, so too will inflation, leading to a sharp selloff in the bond market.
In the two years since Yellen’s Rhode Island speech, the Fed has on every official occasion referred to anecdotes or “signs” of wage increases; but no data. I’ve recounted the actual data too many times, so I won’t do so again here. Needless to write, there are only ever anecdotes and signs even now as the unemployment rate drops to levels that are frankly so far out of line that it can only reveal the further folly of following it.
In terms of inflation, in the last five years, a total of 60 months, the PCE Deflator has registered 2% inflation exactly once. That’s an enormous problem not just because of wages but because of the Fed’s explicit commitment to an official target of 2%. At what point does the world stop believing the Fed can actually hit its target? For both the bond market as well as consumer surveys, the answer appears to have been 2014.
The 5-year/5-year forward inflation rate, already low by virtue of what Stanley Fischer was confessing to, broke down severely in the middle of 2014. This was no small thing, as the forward rate is not simply an indication of market inflation expectations, but rather a perception of the long run implications of what is taking place in the intermediate term. In other words, if inflation is thought to be low today but over time moving back to normal, the 5-year/5-year forward measure will not budge; its concern is only about long-term over the horizon and the state of economic potential.
For the first few years after the Great “Recession”, it moved around in more volatile fashion but anchored mostly to the same level that prevailed before the crisis. What that suggested was grave market doubts as to whether the Fed could actually prevail, economic as well as monetarily speaking. These gyrations were tied to, as you would expect, the QE’s. Each time one was delivered, the market would become enraptured, only to be disappointed until the next one. The events of the “rising dollar” appear to have been the last straw – not because there weren’t going to be any more QE’s but because the prior four didn’t accomplish anything they were supposed to.
For consumer surveys like the University of Michigan’s Index of Consumer Sentiment, though headline “confidence” remains high the responses in terms of long-term (and short-term) inflation are the exact opposite. And like the 5-year/5-year forward rate, they turned lower in the middle of 2014 after being completely unmoved despite the crisis and its aftermath to that point.
None of this matters, however, as conventional wisdom in 2017 is that the economy is improving, inflation is coming, and the bond market is totally wrong (interest rates have nowhere to go but up). It doesn’t matter that, like the 1930’s, the yield curve has flattened, indeed collapsed, and likewise in the wrong direction. Liquidity preferences rule, yet they aren’t today considered just like they weren’t then.
The excuses have become more convoluted as time passes and interest rates don’t, actually, have to rise. One research outfit recently speculated that because central banks around the world have bought up so many bonds the remaining marketable float is dwindled as to be insufficient. Why that wouldn’t spark a 1994-style bond massacre is left unexplained, because if the market truly believes inflation and recovery are coming then a small float would be a big risk in the direction everyone wants.
Earlier this week, Bloomberg published another such apology:
“China and other developing nations are accumulating wealth, but failing to create sophisticated local markets that feature their own risk-free instruments. That’s left a dangerous reliance on U.S. Treasuries, according to Jen’s argument, perpetuating a bond bubble and pushing investors into riskier assets.”
As the article acknowledges, this is a variation of Ben Bernanke’s ridiculous “global savings glut” hypothesis from the last decade that attempted to explain at that time why economists don’t understand the bond market. This version at least acknowledges the correct symptom if the wrong ideas about it. The accumulation of UST’s globally is really the heart of all these issues.
If we are speaking about liquidity preferences in light of the clear and obvious failures of policies and the inability of policymakers to frankly get anything right about the global economy, then holding liquid instruments only makes the most sense. It is the prudent judgment of the market that despite claims of “money printing” there isn’t actually money circulating. We are at Friedman’s point of “tightness.”
That these various places around the world express these liquidity preferences in UST’s follows directly from the thing that is most monetarily lacking – “dollars.”
So it was that this week the Fed “raised rates” for a fourth time and the markets reacted in the opposite fashion. One reason why was the latest release of the CPI published the same day as the FOMC vote. Quite predictably, consumer inflation rates have decelerated due to the almost exhausted boost from nothing but oil price base effects. Oil prices in the space of a few months achieved what five years, 60 months, of monetary policy never could - and then promptly took it away. You don’t think markets noticed that it isn’t wages pushing consumer prices? And what that actually means for more than just oil, the CPI, or even wages?
Bond yields at the long end fell, further flattening the yield curve in the wrong direction, while eurodollar futures were bid suggesting that the future structure of money rates will continue to be very 1930’s. There is no growth opportunity, which is what truly makes the comparison. Wednesday’s vote was, just as the first “rate hike”, another unnecessarily embarrassing day for the Fed.