Volatility Primer: Quarterly versus monthly sales figures

There’s still some data I want to have a look at before posting the April 2010 direct market sales estimates, but from my initial work with the numbers released today by Diamond, it appears that, even against unremarkable 2009 competition, April will be as weak as a month for comics sales as March was a very strong month. As I expect you’ll see other estimates online before our work is complete, this is probably a good time to mention again that, increasingly, the monthly totals are considerably more volatile than the quarterly totals. The chart here shows how the monthly “noise” looks like when superimposed over quarterly unit sales.

The calendar plays a major role. March was a five-ship-week month; April, four. Back in the days when sales charts reported preorders, that fact generally didn’t matter that much, because publishers spread their usual monthly offerings out across the entire month (or skipped the final week altogether). But now the charts are reporting comics that actually shipped, meaning that five-week month has one more ship week in which to capture the premiere of a comic book solicited for sale in a previous month but which shipped late. And, of course, it’s one more week for shipped reorders to fit into the charts.

Over the last six years, retailers have received 6.29 million Top 300 comic books from Diamond in four-week months — and 6.99 million Top 300 comic books in five-week months. Since 2004, final unit orders for Top 300 comics are 11% higher in five-week months. Why not 25%, given the extra week? That’s the effect of the monthly cycle, as mentioned before: All things being equal, most titles are only supposed to have one new issue per month. The extra week is capturing extra reorders and monthly titles which either shipped out of their place in the calendar, or non-monthlies which simply had a greater chance of landing in that month because of the extra week.

So quarterly aggregate totals tend to be considerably less volatile than the monthly totals. But there’s another factor: The third month in a quarter tends to have slightly higher sales than the other two quarters.  Looking at all the completed quarters since 2003 when Diamond began reporting final orders, average unit sales for the first and second months of each quarter were nearly identical. But third-month orders were nearly 2.5% higher. That’s not a lot, and maybe it’s not significant; if you close off the sample to very recent years, the first months of quarters tend to do better. But we can imagine why third months might see more orders: Some publishers may have a fiscal incentive for getting a book to market before a quarter ends; there’s also the fact that the final months of a quarter are more likely to have a fifth ship week than the middle months of the quarter, where February very rarely gets a fifth week. (Although if Diamond closes between Christmas and New Year’s, that may change the calculation some.)

All of this is basically a prescription for looking at wider ranges of time when evaluating market performance. Comics sales charts are monthly for a reason, but sometimes — as I suspect will be the case this month — the longer-term picture will have more to tell us.