NEW YORK (The Deal) -- Evan Schnidman's research goes way beyond the old "briefcase indicator." Schnidman is the founder and chief executive of Prattle Analytics, a consulting firm that uses proprietary data that tells hedge funds, investment banks and other Wall Street institutions how to accurately use Federal Reserve commentary in their decision-making. Despite its name, Prattle probably has a better fix on the Fed than the pundits who based their predictions of Fed actions on the changing heft of former Chairman Alan Greenspan's briefcase. Schnidman is a Harvard Ph.D. and former professor at the school, and the author of the upcoming book, How the Fed Moves Markets. The Deal spoke with Schnidman recently about his research. Must Read: 10 Stocks Billionaire John Paulson Loves The Deal: Explain the origin of the firm's name. Evan Schnidman: "Prattle" is one of those SAT words that maybe one in four people can define. It means gibberish, meaningless chatter. We pride ourselves on cutting through gibberish, and analyzing vital parts of the chatter coming out of central banks in a way that allows us to generate quantitative data. TD: How do you quantify speech? ES: What we do is sentiment analysis. In traditional sentiment analysis, a good buzzword minus a bad buzzword equals sentiment. So it's possible to analyze every word of a comment in the context of what's been said and examine how the market reacted to prior communications. Our algorithm updates constantly. We have a comprehensive metric that looks at every communication that comes out of the central bank. Not just FOMC announcements or the minutes of meetings. We look at speeches. And not just of the Fed as an institution. We also track the speeches and interviews of individual Fed governors. The critical thing is that the data is unbiased. The result is quantitative data that we push out quickly that's comprehensive and unbiased. TD: How do your clients use the data? ES: It's used as a signal in a multifactor model. Most funds invest based on a complex model with many signals. However, to date, they've been forced to analyze central banks on a qualitative model. The qualitative post hoc reading has been the norm. We provide quantitative data on central banks that's unbiased. TD: What's the salient feature? ES: One of the appealing factors is speed. We're able to process any Fed commentary in a matter of seven to 10 seconds. And instead of a methodology that focuses on a single phrase or comment -- typical in qualitative analysis -- we're able to look at the whole of a communication that's been carefully crafted. And it's in a away that's both unbiased and truly comprehensive. TD: So investors have been doing this wrong all along? ES: Eventually they're able to catch on to the correct analysis. But not before they've potentially lost money. There are some key examples in recent memory. In December 2014, the FOMC press release took out the phrase "considerable time." A lot of funds that trade on headlines saw that phrase was taken out and interpreted the communication as being more hawkish than had been anticipated. However, in reality, "considerable time" was replaced with comparable language. Funds that traded on the headline quickly had to rebalance their positions once they realized their error. In another instance, at the end of quantitative easing, Janet Yellen was asked when the Fed would start to raise rates, and she said, "In six months." The market saw that as hawkish. But the larger context of her comments was actually dovish. The result is that the market sold off most of that day and much of the next day, and then spent the following day rebounding. Our data, scoring the press conference comprehensively, got the sentiment right immediately. Must Read: Federal Reserve Drops 'Patient' From Statement, Remains Dovish TD: As a start-up, don't you have a limited sample size? ES: We actually have data going back to the early '90s. And not just for the Fed. We monitor 15 central banks with data going back over 20 years. Though prior to about 1998, the Fed's communications protocol was very different. That's when the "briefcase watch" was the only way to analyze what the Fed was doing. Thanks to transparency reforms, things changed systematically. It began to issue press releases. And that transparency caught on globally. I actually did my Ph.D. thesis on central bank transparency, so this is a subject close to my heart. TD: What's been the reaction among financial institutions to your product? ES: The more algorithmically inclined a firm is, the more quickly they acknowledge the value of the product we're offering. It's another quantitative input that they can integrate to their decision-making. It's a tougher sell with global macro funds that are used to investing qualitatively. But the fact that it's a quantitative product actually makes it more valuable to them, because it acts as a check on their existing methodology. It's another tool that their economists can use. TD: And it applies across asset classes? ES: Our data correlates very nicely with fixed-income investments, which isn't surprising, since the sentiment of central banks correlates with moves in the fixed-income market. The same with the currency market. We have data on both side of a currency paired trade. What was surprising to us was the correlations with the equity market were so close. Which, if you think of it, makes sense, because central banks are a leading indicator of economic news. It's consistently a good buy indicator. But it's an extremely good sell indicator. TD: Does it work as well in periods when central bank actions and commentary are dynamic, as they are now? ES: In fact, our data becomes more valuable because central banks are increasingly using forward guidance as a primary strategy for policy. Look at how they explain policy. Transparency has become part of policy itself. Must Read: Fed Policies Haven't Helped Savers Read more from:
from Latest TSC Headlines http://ift.tt/19AacrA
No comments:
Post a Comment