If the economy is so booming, why did retail sales (ex-autos) decline -0.2% in February? Searching for the usual suspects to blame for unwanted economic data – the weather created a bad hair day, the sun got in consumer’s eyes – Bank of America Merrill Lynch looks at a seasonally consistent phenomenon: tax refunds. Causation for slowing retail sales being based on a one-time factor such as a tax refund results in the bank analysts not to consider this a trend changing event. The impact of online retail sales aside, it is those who are dependent on a tax refund to spend that are the one time factor in weak retail sales.

Large swings in retail sales numbers have resulted in a slightly lower average

Retail sales have exhibited sharp price volatility over the past few months. Technical market watchers sometimes attribute volatility to an underlying disruption in the forces of supply and demand. Certain professional traders claim to have identified idiosyncratic stock volatility prior to the 2001 and 2008 market crashes, for instance, that were execution triggers to adjust exposure factors.

BofA’s Chief Investment Strategist Michael Hartnett, US Economist Michelle Meyer and the macro analytics team — which includes “data driven” quant Savita Subramanian – note a trend change could be occurring in retail sales and want to understand the recent larger than average decline.

Measuring a three-month moving average of retail sales numbers, they note a slowing trend line. Retail sales have been increasing 0.1% on average over the past three months, less than the previous trend rolling average. February’s beyond mean adjustment, down -0.2%, was offset by equally strong numbers previously.

A moving average can belie the actual volatility in the retail sales number, which is the case here, the report notes. There were large swings in retail spending over the prior two months three-month month average masks. This is an example of two points of noncorrelated volatility – in this case retail sales numbers went in different directions in January and February — combining to deliver dampened volatility through an average.

Considering patterns of behavior that deviate from the past is, at times, is a method to understand volatility. Here the “BofA analysis” – evidently eschewing Merrill Lynch in the title – discovers that tax refunds have been delayed and this unique event is impacting performance. Average tax payers, eager to get their hands on an above mean annual tax return bounty, are waiting to spend due to a change in tax law that delays refunds to early filers. This rather average concern is causation for the below average February retail sales numbers.

Credit card and general debt experience relative value divergence

Relative value analysis is often a method for quants such as Subramanian to measure trend strength. Here the report compares general debit and credit card spending relative to a baseline norm.

Debt and credit card spending is an is an extension of a household budget, an indication of “dry powder” available in household spending. The BofA report considers it a “good indication since presumably usage of debit cards should be more sensitive to the tax refund (proxy for cash) than credit cards (leverage).” BofA analysis discovered that retail sales ex-autos for debit cards declined 1.7% month over month while credit card spending was up 1.8%, a divergent path.

Those who plan spending based on a tax refund are more likely to be lower income households, as they must make decisions without much saved in reserve. “We see this clearly in our data where the lowest income quintile reduced spending by 3.4% while the highest income quintile actually increased spending by 0.9% (month over month),” the report observed.

Given these one-off adjustments, BofA thinks the move in retail sales is likewise a one-time event due to the tax refund. “It makes sense to discount some of the weak signal,” the report said. “We think it is likely that we will see a bounce back in March or April as the tax refund is spent later in the spring.”