Those of you in Boston, we visited accidentally this week.

Flying into LaGuardia Sunday afternoon, Mother Nature had thrown up a snowy blockade, and running low on fuel after an hour circling like a speedway pace car, we diverted to Logan. Thanks for the gas.

We caught our breath and de-iced, and the captain came on: “Ladies and gentlemen, I don’t know how this is going to go”—pilots should never start with those nine words—“but we’ll be taking off shortly. We may just circle New York and be back at Logan. Who knows?  Thanks for flying United.”

Without offense to the great state of Massachusetts the decaying glory of LaGuardia drawing nearer filled us with thankful anticipation. All’s well that ends well. But we’d assumed we’d fare better with a March New York sojourn than a January one – an assumption lacking buttressing data.

Weather may be the exception, but as a rule, challenging assumptions is an invigorating intellectual process. In his 2003 book Moneyball, Michael Lewis tells how Oakland Athletics manager Billy Beane assailed the battlements of baseball. He tested the game’s assumptions about winning by applying data to conventional wisdom.

That refreshing opportunity prevails in our profession daily. Fourteen years after I saw my first Bloomberg screenshot as an investor-relations officer under the heading “you broke through your moving averages” after I asked my listing exchange, “Why is my stock down today?” our clients still get the same answer I did.

No doubt some traders use moving averages. But the bulk of money moving markets today follows asset-allocation models as the dominance of Blackrock and Vanguard illustrate. Data from the Investment Company Institute show that over 90% of assets in 401k plans now use some sort of asset-allocation model.  Do you think these track moving averages?

Other assumed IR wisdoms unsupported by data:

• Short interest is the best way to measure risk in shares. 
• Big volumes mean big investors are buying or selling.
• If my shares behave differently than my peer’s, someone sees them differently.
• Tracking share-ownership is an effective way to understand your stock’s activity.
• Ignore high-speed trading because it’s noise – and in fact, ignore the stock.

Have you graphed short interest and price behavior to study correlations? We have. There is no consistent predictive quality to short-interest levels. So we use other measures of risk that offer statistically significant predictive characteristics.

We routinely measure high volumes having an absence of substance. High-speed traders and derivatives-pairs both foster volume inflation. And ETFs – real money yes, but money that can change its mind daily – comes and goes in block trades routinely.

Price-movement in your shares is a product of supply and demand more than story. If an algorithm aims to buy the same number of your shares and a peer’s and your peer’s shares are readily available and yours are not, your price will soar and your peer’s will not move (this is why understanding supply and demand in your shares is vital).

If 43% of market volume is borrowed, and ETFs post positions every day, and indexes are continuously adjusting to track the measures they’re paid to peg, how could settlement data removed T+3 from action provide reliable data on price-changes?

Investor-relations practitioners remain so far as I know the only product managers instructed by somber-faced docents to disregard the data on pricing, distribution channels and other key factors essential to performance in any other product-marketing endeavor.  The single greatest marketplace sensor feeding data back to you about drivers in your market is demographic price-setting data – especially high-speed traders. They’re intermediaries at the nexus of supply and demand.

When the intermediary was the specialist on the floor, he or she was your best feedback mechanism. Today with three of four designated market-makers high-speed traders, the feedback is just as important. But we measure it now with machines rather than humans.

Here’s your challenge for the quarter. Find a couple assumptions you’re making about the IR job and challenge them with data.  You might find better ways to do things. Also, always check the weather report (they’re right more than wrong and that’s statistically meaningful!).