Thursday, June 02, 2011

Even more on news driven trading

News driven trading is even more in vogue today than when I last mentioned it, judging from the increasing number of vendors (e.g. Ravenpack, Sensobeat, Recorded Future, etc.) and researchers pitching their wares. Not only are traditional financial and economic news deemed important, but researchers have found even blog posts (at least those on Seeking Alpha) and Twitter (Hat tip: Satya and William) to be predictive of stock prices.

One key ingredient to success in this type of trading is of course the ability to gain access to breaking news ahead of other traders. On the macroeconomic news front, the MIT Billion Prices project has spun off a company called PriceStats to deliver daily consumer product price index to subscribers. PriceStats compiles this index by continuously scanning online retailers' websites, and hopefully provides a preview of the official CPI numbers. Whether this is useful for futures and currencies traders is of course subject to their rigorous backtests, though the chart displayed on their website does suggest that the daily price index is a leading indicator of the CPI.

There is an important caveat to using news trading: not all news are equal. So another key ingredient to success is to carefully differentiate between the different types of news and backtest their predictive abilities separately. For example, I recall some research has indicated that an analyst downgrade of a stock from a "hold" to a "sell" rating has more impact than from "buy" to "hold" rating.

My own experience with news driven trading is that for all this trouble, the trading opportunities are relatively few compared to pure price driven trading, the consistency of success is low, and finally the profitability lifespan is short. If you have better experience, do share it with us.

10 comments:

robert said...

re Twitter predictions. The article is a scam and the HF itself is a scam too.

An exemple:Derwent has previously claimed that its model is 87.6% accurate. The firm said it is targeting returns of between 15% and 20%.

87.6% daily accuracy on the DJIA would translate to (2*0.876-1)*sqrt(2/pi)*sigma=73bps/day with sigma=20%/sqrt(252)... far from the 15-20% announced...

Another example. The 86.7% announced has to be compared with an almost 80% accuracy using past DJIA returns only. On top of that, out-of-sample testing is on a 15 days period!

Ernie Chan said...

Robert,
I agree with your comments fully.
The major problem with twitter prediction is the lack of a long, real trading track record.

I can't believe people would invest $100M into a fund without that kind of track record. But I guess in a tulip mania, everything is possible.

Ernie

Anonymous said...

Great post!

Do you know of any available applications that incorporate Twitter or other real-time search tools?

Proprietary?

Open-source?

Ernie Chan said...

Anon,
I am not aware of open-source news search applications, but as I mentioned in my post, Ravenpack, Sensobeat, Recorded Future are 3 of the vendors in this field.
Ernie

Slacker said...

I trade news based strategies for a living, but take an entirely different approach. Rather than try to beat everyone to the punch I enter trades on the opening cross of the first session following the news report. It's archaic from a quantitative standpoint, but it requires nothing more than Excel, a little VB programming and Yahoo Finance. It's also very reliable from an implementation standpoint, i.e. slippage is essentially zero. PEAD is one such strategy.

Ernie Chan said...

Slacker,
Yes, I used to trade PEAD too. But the returns became too erratic, and perhaps the time horizon to exit became too short, for me to continue.

Good to hear it is still working for you!
Ernie

Slacker said...

Ernie,
When were your entry points relative to the earnings announcement? I have found the day immediately following the announcement needs to be treated separately from subsequent time periods.

What type of exit criteria did you use? Analyses I have run using tick data have shown that entering on open and exiting anytime after 3:00 provides the optimal return for that day.

Any attempts to optimize my stop loss criteria, however, have simply resulted in lower expected returns. Even for multiple day time frames, "no stops" works best despite my best efforts to incorporate price, volume, trailing stops, true range or any other indicators in the logic. I end up using 10% hard stops as an aribtrary "knee in the diode curve" to mitigate risk, but this remains a vexing issue nonetheless.

Have you tried developing news based strategies on other asset classes such as commodities or currencies?

Ernie Chan said...

Slacker,
I entered and exited at the same times you mentioned. I also find that stop loss is not a good idea.

I have not traded news on other instruments, and have turned my attention to purely technical indicators for both reversal and momentum strategies.

Ernie

Anonymous said...

Hi Ernie,

Another way to trade the news without racing the big guys is as an input to a mean reversion or momentum strat. There is academic research supporting this (Wesley Chan and Tetlock). I have done some simple confirming analysis with daily stock returns and yahoo news items but not actually traded it. Basically a large formation period return with news is different than one without news - more tendency to mean revert without news, more tendency to trend with news.

Ernie Chan said...

Hi Anon,
I agree with your observation: news is the driver of trending behaviour. Stock price typically mean-revert in the absence of news.
Ernie