T-Minus One Day: How Were The Markets Positioned The Day Before The Last Three Rate Hikes

Nine years after the Fed's last rate hike on Jun 29, 2006, the 17th and final consecutive 25 bps rate increase in a tightening cycle that started in June of 2004, we are now at D-day, with just over 24 hours until the Fed's alleged first rate hike announcement in nearly a decade.

So as the Fed huddles in the Marriner Eccles building for its latest "most anticipated ever" meeting at least until the next one, looking at such deplorable economic data such as the first headline deflationary CPI print since January, following a series of disappointing manufacturing, housing and retail sales reports not to mention a payroll number that absent sugarcoating was a solid miss as well, not to mention global emerging markets that one after another are sinking into recession and exporting record amounts of deflation to the US, one wonders: what happened the last three times the Fed launched tightening cycles in 2004, 1999 and 1994? And more importantly, what did the market think was going to happen?

First, here is what the Fed Funds futures believe is the probability of a rate hike tomorrow: as of moments ago, it was 30%, down a fraction following the CPI report.

So what about in prior years?

As Morgan Stanley points out in a report by Guneet Dhingra, both 2004 and 1999 were perfectly expected, at least as far as the market is concerned, rate hike cycles.

In the 1999 hike cycle, the Fed hiked 25bp, or 50bp, or stayed on hold when these outcomes carried at least a 50% probability in market pricing, on the day before the announcement (see Exhibit 2). This means that the Fed did not deliver significant surprises in its 1999 hike cycle. For example, in October and December 1999, the market implied probabilities for an ‘on hold’ outcome were well above 50% on the day before those FOMC meetings (see Exhibit 2). Similarly, the other outlier outcome - the 50bp hike in May 2000, was also priced in with a 50% probability on the day before the announcement.


The 2004 hike cycle was even more ‘predictable’ as the market priced in at least a two-thirds probability of +25bp hikes at least four weeks before the FOMC, and close to 100% probabilities on the day before the FOMC meetings (see Exhibit 3). This was a result of the ‘measured pace’ language that the Fed used in the 2004 cycle.

A summary of the implied probabilities stretching from 2 months to 24 hours before any given rate hike in the 1999 and 2004 tightening cycles:

In other words, the lastt two rate hike cycles were very much anticipated by the market and did not come as a surprise. Where it gets interesting is looking at what happened in 1994.

As Morgan Stanley reminds us, and the 50% or so of current Wall Street workers who were not active back then, "The 1994 hike cycle was one in which the Fed did not sufficiently manage market expectations prior to liftoff. This resulted in a February 1994 rate hike that was followed by high levels of market volatility and a significant tightening in financial conditions."

So what happened in 1994 and why does Morgan Stanley think the Fed has "learned its lesson from 1994, and the significant tightening in financial conditions that occurred that year."

The 1994 hike cycle – while broadly anticipated – surprised investors in terms of the timing and magnitude of rate hikes along the way. With an improving economy in 1993, the stage had been set for a rise in interest rates. However, as Exhibit 4 shows, the Fed surprised the market many times, including with its first rate hike in February 1994. Going into that hike, the market had only priced in a ~22% probability of liftoff one week before the February 1994 FOMC meeting.

A quick primer on why the Fed hates surprises: "Surprises matter to the Fed, because surprises can lead to a bigger tightening in financial conditions than is warranted." I.e., the market will dump.

And yet, this is where Morgan Stanley is way off, because while it is correct that 1 week before the rate hike, the Fed Funds future implied rate hike probability was just 22%, 5 weeks prior it was 53% and 6 weeks prior it was 61%.

But most importantly, the day before the February 1994 rate hike, the probability had soared to a whopping 70%, and it only went higher from there for the March and April rate hikes as can be seen in the table below.

In other words, as much as Morgan Stanley wants to pass of the 1994 rate hike cycle as a surprise, or a mistake, it was anything but with implied rate hike probabilities all higher than 70% the day before the rate hike!

What about the "2015 rate hike cycle"? As noted up top, as of this moment, 1 day before the alleged first rate hike in 9 years, the probability is a paltry 30%.

So if the 1994 market reaction - especially in the bond market - was one of shock and awe when the market was quite prepared ahead of the actual rate hike announcement, what would happen tomorrow as the probability is more than 50% less than it was in the "turbulent" 1994.

All that said, this analysis ignores the all too critical point of perception reflexivity, and the reality that the market volatility that surges as the rate hike day inevitably approaches is precisely the reason why the market is projecting such a low probability to a rate hike: the Pavlovian dogs trading various asset classes are so trained to expect perpetual easy money conditions from the Fed, they will do anything, certainly keep rate hike odds as low as possible, to deter Yellen from at least attempting monetary normalization and raising the cost of money for the first time in nearly a decade.

So that is what the markets thought just before the last three rate hikes. As for what Janet Yellen is thinking, tune in in just over 24 hours for the answer although now that even Goldman's CEO Lloyd Blankfein has spoken, saying earlier that "the data does not seem to support a interest rate increase this week", tomorrow's outcome is all but decided.