I’m sure there is a textbook definition, perhaps even clinical, of recency bias, but to me the term refers to an inability to see past the margins. In short, it is a failure of imagination. We see this most clearly in econometrics and statistical analysis, where “tail risks” happen far more frequently than the numbers predict. Rather than rethink the theories behind the models and statistics, it’s easier to cling to standard deviations and modeled assumptions even when the statistically “impossible” is staring us in the face.
The most obvious example of recency bias was Chairman Bernanke’s (then just a Governor and voting FOMC member) reasoning against a housing crash in July 2005:
“It’s a pretty unlikely possibility. We’ve never had a decline in house prices on a nationwide basis.”
At each and every inflection point in economic history, the full weight of the mainstream is reduced to grasping at the recent past in full expectation of future similarity; rationalizations being given more consideration than careful examination.
On September 5, 1929, Irving Fisher, perhaps the most famed economist of the age, was quoted in the New York Times proclaiming, “There may be a recession in stock prices, but not anything in the nature of a crash.” Unfortunately, Fisher redoubled that thought, with more emphasis, in his infamous proclamation of October 17, 1929:
“Stock prices have reached what looks like a permanently high plateau…I expect to see the stock market a good deal higher within a few months.”
Fisher was not alone in his optimism in due course of one of the most obvious asset bubbles in history. In Business Week on November 2, 1929, the bubble was acknowledged and then casually dismissed:
“For six years American business has been diverting a substantial part of its attention, its energies and its resources on the speculative game… Now that irrelevant, alien and hazardous adventure is over. Business has come home again, back to its job, providentially unscathed, sound in wind and limb, financially stronger than ever before.”
To put a fine point on this line of inquiry, Treasury Secretary Andrew Mellon rationalized in December 1929 his course of optimism even in the face of that economic and financial inflection, “I have every confidence that there will be a revival of activity in the spring, and that during this coming year the country will make steady progress.” That is a sentiment that has been echoed and revisited every single year since 2009. Perhaps when a current FOMC official expresses the same confidence there is more weight in an age of dour economic management through central banking, but results give no reason to support such a distinction.
The CBO is a creature of econometrics, feeding an endless stream of supposedly objective assumptions into its elegant statistical models to produce “predictions”. In its January 2007 outlook, the CBO saw the slump in housing, like Bernanke, as a slight bump in the economic road, “CBO forecasts that GDP will grow by 2.3 percent in real terms in calendar year 2007 but by 3.0 percent in 2008.”
Even after the eurodollar crisis in August 2007, followed closely by the liquidity impacts on bank financials, the CBO in January 2008 was still caught up in recency bias that infected every mainstream economic strain:
“The pace of economic growth slowed in 2007, and there are strong indications that it will slacken further in 2008. In CBO’s view, the ongoing problems in the housing and financial markets and the high price of oil will curb spending by households and businesses this year and trim the growth of GDP. Although recent data suggest that the probability of a recession in 2008 has increased, CBO does not expect the slowdown in economic growth to be large enough to register as a recession.”
There are no predictive abilities in any of these models outside of the margin of error. Once events stray from predicted trajectories, these “predictive” capabilities become reactionary rationalizations. The CBO expected 3% growth for 2008, but the slowdown in 2007 was worse than feared, so the changes to 2008 expectations were driven by events after they happened. The CBO and its models were not leading, they were behind the curve the entire period trying to catch up and maintain their modeled sense of order and meaning. They were incapable, by definition, of seeing the world outside of that order, even as events were demonstrating it conclusively in real time.
The same has been, and still is, true of all economic models. The Federal Reserve itself has laid out just how bad its predictive powers were in the face of what was a “near impossible” course of events. It was assumed (by assumed I mean modeled) at the end of 2007, already into the real events that were destroying liquidity and economic function, that the probability of a zero federal funds rate (ZLB) occurring before the end of 2012 was between 2% and 9%. The probability of an eight-quarter ZLB was “trivial” across all six of its main modeling systems.
Even after incorporating the data and empirical reality of what actually did occur in 2008 and 2009, only the GARCH model saw an eight-quarter ZLB at greater than a 1% chance (still at only 3%). The paper does not specify what the assumed probability of a now eighteen quarter ZLB works out as, but I can safely “assume” it was something very close to “impossible”. That means these models have a very different definition of what impossible actually means, and that is what recency bias and the failure of imagination actually is.
These are not exclusively academic concerns, as they are embedded in business forecasting and even the assumptions and analysis of the economic accounts themselves. If the GARCH, ferbus, EDO and TVP-VAR models all concurrently assume that recession is “impossible” and growth is “just around the corner”, then the BLS, BEA and Census Bureau add those premises to their benchmark and adjusted assumptions contained within the economic accounts themselves; it comes out in the data we use to measure exactly where we are in a “cycle”. If the predictive models say that recession is unlikely, then it gets put into the economic account assumptions that show recession as unlikely, which leads to even more predictions of low probabilities of recession and leads to further statistical “confirmation” on the same basis. Such circular logic and feedback explains much here.
Because statistical models are captured by their empirical data set, no matter how sophisticated and elegant they incorporate “jump models” and expectations they will never be able to see inflections beyond the data. And one of the most potent pieces of recency bias is the assumed efficacy of monetary “stimulus”. Models see applications of QE as equivalent to actual growth in the future, so they simply and blindly assume growth whenever QE lands. It explains a good portion of why we see so much dichotomy and volatility in data series and economic accounts.
In that very real sense, they are only placing themselves in a long and dubious line of mainstream projections and rationalizations that refuse to acknowledge the true complexity and uncertainty in the real world – even when it is staring them square in the face.
Click here to sign up for our free weekly e-newsletter.
“Wealth preservation and accumulation through thoughtful investing.”
For information on Alhambra Investment Partners’ money management services and global portfolio approach to capital preservation, contact us at: jhudak@4kb.d43.myftpupload.com
Stay In Touch