A Blog by Jonathan Low

 

Jan 9, 2017

How Credit Card Data Might Be Creating Stock Trading Distortions

A reminder that 'predictive' is a relative term. JL

Miriam Gottfried reports in the Wall Street Journal:

Investors buy data on credit- and debit-card transactions to gain an edge ahead of earnings news. Data tracks how consumers use apps on mobile devices, receipts culled from email inboxes from users who accepted terms of service of certain apps (and) transactions on noncarrier-owned mobile-phone stores. That has added volatility to stocks. The lesson may be not to get lost in the data. Trying to predict company earnings has been a loser’s game.
Credit-card data sold to investors is making shares of retailers behave strangely, especially when the data gets things wrong.
Investors in retail stocks used to count cars in parking lots, track foot traffic in malls and look for discounts on the racks. Today, they buy data on credit- and debit-card transactions to gain an edge ahead of earnings news.
That has added volatility to the stocks. When Tailored Brands, the owner of Men’s Wearhouse and Jos. A. Bank, reported mixed earnings and same-store sales growth late on Dec. 7, its shares shot up nearly 40% the next day. A big reason for the jump was that short interest—or bets against the shares—had climbed to 20.1% of the stock’s float by Nov. 30, up from 13.7% on July 30, the day of its second-quarter report, according to FactSet. Those investors appear to have been following card-spending data supplied by firms such as Earnest Research, which showed the business was performing much worse than expected. So many investors appeared to be betting the shares would fall that when they tried to close out their short positions by buying shares of Tailored Brands, the stock soared.
Earnest Research is part of a new breed of companies that sell credit- and debit-card data to a number of hedge funds and other big investors. The data’s prevalence, and its predictive power, suggest it is increasingly affecting investment decisions. Trading in shares of retailers like Tailored seems to back that up.
Envestnet’s Yodlee division is a leading provider of the data. Yodlee’s primary business is selling online personal-finance tools to banks and other financial institutions. But it also anonymizes the data it gathers from credit- and debit-card transactions and sells some of it directly to hedge funds and to research firms such as Earnest Research, 1010data and Second Measure. The firms then crunch the data, package it and sell it to investors, sometimes incorporating other data sets or mixing in their own analysis and commentary. They typically don’t provide “buy” or “sell” recommendations, and their data represent only a subset of transactions at a given merchant. Two days before Tailored Brands’ shares jumped, the stock of Francesca’s Holdings, a women’s clothing retailer, climbed nearly 29% when it announced strong earnings. Subscribers to Earnest Research had seen numbers suggesting the quarter was good, but the data, which is released every other week, showed sales down in the first 23 days of the fourth quarter. Negative bets on the stock rose. Then, Francesca’s raised its earnings guidance instead of lowering it as expected, forcing investors to cover their short bets and sending the stock up.
Volatility in retail stocks has increased versus the broader market over the past year and a half. At the end of June 2015, the S&P Retail Select Index was 1.2 times more volatile than the S&P 500. By the end of 2016, it was nearly 1.5 times more volatile, according to a Wall Street Journal analysis of FactSet data. Reasons for that could include retailers’ struggles with managing physical stores as more sales move online, but the wider use of credit-card data may also be playing a role.
Investment firms can pay millions of dollars for the data, and in many cases it points in the right direction. Subscribers to 1010data saw that sales were pacing down 15% for athletic retailer Finish Line in the third quarter. The shares fell 8.7% after its Dec. 21 earnings report. And 1010data also signaled lower-than-expected quarterly sales for Bed Bath & Beyond, whose shares fell 9% on Dec. 22, the day after its weak report.The data’s influence doesn’t end with retailers. On Oct. 5, Earnest sent a note to clients saying Netflix’s paid domestic streaming subscriber numbers were tracking below consensus at the end of the third quarter. Short interest climbed leading into the earnings report, and the shares shot up 19% on Oct. 18 after Netflix beat consensus estimates on that figure.
There is lots of other data that buy-side investors purchase to help them make decisions. Research firm 7Park Data, for example, tracks how consumers use apps on mobile devices. It is also one of a number of research firms, including Quandl and Leucadia National’s MScience, that license anonymized data on receipts culled from email inboxes from users who have opted in by accepting the terms of service of certain apps. In addition to credit-card data, MScience also licenses anonymized transaction-level data on more than 10,000 noncarrier-owned mobile-phone stores across the U.S.
For investors, the lesson may be not to get too lost in the data. Trying to predict company earnings has usually been a loser’s game. The big price swings in retail stocks make it a more dangerous one.

0 comments:

Post a Comment