S&P posted wrong number for 12-month return, Minneapolis firm caught it
- Blog Post by: Adam Belz
- July 10, 2013 - 11:31 AM
A 3 percent difference in the 12-month return on the S&P 500 represents more than a few pennies.
That was the size of an error reported by the vaunted stock index for a few days after the end of June, and flagged by a firm in Minneapolis (among others) before it was corrected.
S&P Dow Jones reported at the end of June that the 12-month total return on the S&P 500 was 23.6 percent. Only problem was, the real return was 20.6 percent. The number, now corrected without apparently doing any harm, is used by portfolio managers across the country to gauge their performance and explain their progress to customers.
Here’s how the storm in a teapot happened, according to Howard Silverblatt, a senior index analyst for S&P Dow Jones Indices: June 28 was the last day of trading in June 2013, and to calculate the 12-month return on the index, Silverblatt’s analysts compared June 28, 2013 to June 28, 2012.
Logical? Yes. A figure that reflects the 12-month return on the index? No.
The last day of trading in June 2012 was the 29th, and it was a huge day for stock markets.
“The killer here...is that June 29, 2012, was a great day, up 2.49 percent,” Silverblatt said. “It was a very big day in the market and that made a big difference in the return. If it was a normal day, it could have gone on and noone would have noticed it.”
One person who noticed it was Tony Shink, a director at Minneapolis Portfolio Management Group. (S&P said MPMG was not the first firm to notice the error.)
“I was unable to match the performance numbers after trying to calculate them a couple of ways,” he said.
He thought S&P had perhaps changed components of the index and called them on July 2. When he still couldn’t make the numbers work after a chat with an analyst, he and Kathryn Hubert, a new graduate of the Carlson School who just started at MPMG, called again on July 3.
“When we called the S&P, they were like, ‘Um, no, our calculations are correct,’ and then we go, well, can you please recheck June?” Hubert said. “He said, ‘Hold on, hold on, you’re right. We have this under control, we’ll get this changed right away.’”
MPMG was only partially right, however. Yes there was an error, but their theory for why the mistake happened – they thought S&P had put a plus sign in front of the June return instead of the accurate negative June return – was not correct.
Ultimately, the dust-up is not all that consequential, said Matt Paschke, a portfolio manager for Leuthold Weeden Capital Management, because most traders work with the S&P’s raw data to come up with the 12-month return anyway.
“Most people calculate it on their own,” Paschke said. “Anybody putting the two dates into their system is going to use June 30, 2012.”
Some people might chuckle to hear the S&P had a screw-up, but that’s it, said Paschke.
“Seems like a fairly non-consequential error,” Paschke said. “Everybody will just correct it.”
Silverblatt said he apologizes to anyone who saw the wrong number, which is, as far as I can tell, not many people.
“It’s good that it was found,” he said. “It’s bad that it happened.”
© 2017 Star Tribune