And so the great American retail recovery continues being delayed following the third consecutive miss in headline retail sales in a row, despite what was ridiculous and erroneously touted as a record spending season. Well here come the 'product returns' as retailer margins shrink further into negative territory. Today's advance retail sales number came at 0.4% on expectations of a 0.8% increase, from a downward revised 0.0% (0.1% previously), which makes a mockery of both the car sales numbers in December which were the weakest link in today's retail sales, and of surging consumer credit as it proves beyond a shadow of a doubt that US consumers are now using credit cards for the most basic of staples, forget discretionary purchases! And while the number below the headlines was modestly better with ex autos and gas coming at 0.6% on expectations of 0.5%, the prior revision took December to a decline -0.2% from a previously unchanged number. In other words, expect today's ex cars and gas number to be revised to a miss next month as as the Census Bureau learns some key number fudging lessons from the BLS. Yet here is the punchline: there have not been three consecutive retail sales misses since... drumroll please... July 2008. And we all know what happened in the months following. For those who don't, here it is.[Chart follows]

The topic of BLS propaganda seasonal adjustments has been discussed extensively here especially in light of January's NFP beat. We'll leave it at that. However, we were rather surprised to note that the Census Bureau may have also ramped up its seasonal adjustment "fudge factoring" because when looking at the January headline retail sales data, which naturally was a smoothly continuous line on a Seasonally Adjusted basis, rising from $399.9 billion in December to $401.4 billion in January, something rather odd happened in the Unadjusted data set: the plunge from $459.8 billion in December to $361.4 billion in January, or -$98.5 billion in one month, was the biggest one month drop in retail sales in history. Now we won't say much on this topic, suffice to say that it would be far more useful if the BLS and Census Bureaus were to open up their models and explain in nuanced detail just what "old normal" adjustments they still incorporate into data sets. Because as many have already noted, seasonal adjustments used for data from 1980 to 2008 when "up" was the only allowed direction for everything, are completely irrelevant and misleading in the New Deleveraging Normal. Which reminds us: Zero Hedge will offer $10,000 to the first BLS employee to share with us the full and complete excel model set, including assumptions, data tables, and comprehensive output parameters that the agency uses to go from input A to output X. We hope that by spending that money we will finally do society a service and open up to everyone just how it is that the BLS adjusts its Non-Farm Payrolls data.[Chart follows]