For almost 20 years now that return number has gotten smaller. Although we've bounced off our low of 8.68 percent (in 2008), the downward trend returned last year. From 1928 through the end of 2015, the average annual S&P 500 return stands at 9.50 percent.

Allow me to delve in the mystique of perception psychology. Analysts tend to avoid the subject due to its “hocus-pocus” sounding name. Nonetheless, behavioral anomalies are real.

More than 10 years ago I referred to this phenomenon as the “snapshot-in-time” anomaly. It's an effective graphical representation of the “recency” bias. It occurs when you place more weight on something you've experienced more recently. It explains why so many people think Tom Brady is a better quarterback than Johnny Unitas. We are more naturally disposed to emphasize the near-term at the expense of the distant past.

We are ripe to be duped by the “snapshot-in-time” anomaly. Being already predisposed with a confirmation bias, we look for data consistent with our meme of “the economy stinks and it'll never get better.” We quickly find it in the figures presented earlier.

But what if we consciously seek to thwart the “snapshot-in-time” anomaly? What if we look deeper into the internals of the historic numbers and apply them more precisely to the real world of long-term retirement projections? The fact is, many people are looking at uninterrupted saving (and investing) within a 30- to 40-year time frame. If we look at all rolling periods within that range of years, we see the median annual return is roughly 11 percent. Even accounting for the sequence of return risk shows that only of those rolling periods experienced a return (barely) below 9 percent. This low point was the result of the “snapshot-in-time” effect.

Which way should you go? I'd suggest the past is prologue, but no one listens to me. But, maybe this time it's different.

(Figures from the Stern NYU annual S&P 500 return data base.)