* Investors look to gauge sentiment in choppy markets
* Data from Twitter, Facebook used in investment models
* Traditional valuations have become harder to trust
* Sentiment key in markets driven by policy action
LONDON, Oct 30 (Reuters) - After a year in which investors have sometimes appeared to change their view of an entire market on a whim, gauges of market mood are gaining prominence alongside more traditional measures of stock price valuation.
The debt crisis which began in 2007, the subsequent slide in equities and policy-fuelled attempts to resuscitate flatlining markets have prompted investors to add a new breed of sentiment indicators to the mass of data on which they base decisions.
Analysts say sentiment has been a major driver of the periodic wholesale rush into or out of equity markets - charactersised in the jargon as risk-on or risk-off - as investor confidence in policymaker efforts to tackle the global credit crisis has ebbed and flowed.
``In the current market cycles, we have noticed that the performance has been primarily driven by sentiment,'' said Ankit Gheedia, an equity and derivatives strategist at BNP Paribas.
Approaches to measuring sentiment are numerous and varied.
Hedge fund Alphagenius, for example, takes data from Twitter and social internet sites like Facebook, analysing millions of blogs, tweets, forums and news sites for positive and negative comments. It then uses artificial intelligence to develop investment models.
By trying to understand underlying potential swings in market mood, investors can anticipate stock market moves that may not be explained by traditional valuations and corporate fundamentals, such as balance sheet quality.
For instance, compared with historical levels of around 14 times, European companies have been undervalued on a price-to-earnings basis for the last six years. While this partly reflects the harsh macroeconomic backdrop, analysts say it highlights investors' gloomy view of stock valuations.
BNP's Love-Panic index uses a range of data, from fund flows to earnings revisions, to measure market sentiment and has been a fairly accurate gauge of equities' expected returns to investors over a six month period since 2007.
When the aggregate of these indicators reaches very negative levels (panic), it is time to buy. When investors are in love with the market, it is time to sell.
BNP's data showed if investors had bought European equities when sentiment troughed in May 2009, they could have pocketed a return of around 60 percent. The model currently signals an average return of 5 percent in the next six months.
``Understanding sentiment has become quite important to predict prices considering the big monthly moves that we have seen recently,'' BNP's Gheedia said.
To the dismay of euro zone politicians, who have accused markets of irrational or excessive reactions, markets have delivered swift verdicts on efforts to curb the debt crisis.
As a result, European equities endured a rollercoaster 2012, with steep rises and falls in the first half of the year as sentiment on these efforts fluctuated. This was followed by a sharp rally as central banks stepped in to prevent a collapse of the financial system.
``Because we have had so many massive policy changes in the last few years, understanding financial market moves has depended a lot more on the analysis of sentiment than it has the real economy, because the real economy lags the reaction to the financial markets,'' Michael Metcalfe, head of global macro strategy, at State Street Global Markets said.
State Street has its own investor confidence index, based on the buying and selling patterns of institutional investors.
While nimble hedge funds such as Alphagenius use sentiment in their investment models, advocates of the new indicators say they can smooth out the volatility that often follows policy announcements.
``A small group of very sophisticated shops whose influence in the market is strong but limited to minutes and possibly hours, account for the high volatility observed pre- and post these regular announcements,'' said Armando Gonzalez, President & CEO of RavenPack, which provides research based on analysis of news and other data.
While these trades may not be responsible for the overall direction of markets in the longer term, those investors who can get ahead of these volatile movements with ``smarter'' quantitative models that incorporate multiple factors in their decision-making are more likely to win in the longer term.