"The markets crashed on rising crude oil prices."
"Banking, Realty stocks hit on high inflation data."
"Markets bounce back on the back of prospects that the UPA government will survive its full term".
You would have heard or read such comments quite frequently on business channels and in newspapers. The media tries to explain every market move. It tries to find reasons for changes in price. So if stocks go down and crude oil prices also have also risen, they simply put together a causation: Rise in crude prices caused stock prices to fall.
(These days there is a popular view that stock are falling primarily because crude oil prices are so high. Absent the rise and stocks would rise, is the implicit logic. I think this is just an excuse.)
I have often noticed how this explanation is far removed from what happens on floor of the stock markets. Sometime ago, on a particular friday, inflation figures came our slightly higher than expectations, but not too divergent from the expected figure. The markets promptly went up with the Nifty gaining 20 points. The markets stayed high for a few more hours but eventually collapsed at the end of the day to close in the negative. The newspaper headlines next day screamed how markets were spooked by high inflation.
The truth was that markets actually went up after the news and stayed there for quite some time. Clearly, the inflation figure did not cause markets to fall. What caused them to collapse? - we don't know for sure. All we know is that the prices went up when inflation figures were released and they went down at the end of the trading day.
According to the media, a $2 rise in crude oil prices caused stock markets to go down on a particular day. However, on another day, a $5 rise in crude oil prices saw the stock markets going up without any other positive news. So why did a greater rise in oil prices not cause markets to go down with even greater force? We don't know!
Media will keep searching for reasons behind price movements. It simply draws cause and effect relationships with an external event and market moves. However this might just be a convinient excuse to explain market phenomenon. As we have seen, the same external event causes diametrically opposite moves in the markets, across time. Sometimes markets fall after bad news. Sometimes markets rise after receiving the same bad news. The fact is that we really do not know why markets move on any given day. The only thing we can say is that if selling is greater than buying, prices fall. If buying is more than selling, prices rise. That is the most obvious thing there is, but in reality that is all that we know about price movements.
So the headlines should be like:
"Markets went down as selling exceeded buying. Reasons unknown!"
I doubt we would ever see such a headline. For us, what matters however is whether prices are up or down. True reasons perhaps we shall never know...and don't even need to know.
Subscribe to:
Post Comments (Atom)
2 comments:
Yes u hv a point. I have a question though not too related to ur blog here, but let me ask anyways - I know u look at historic data (a lot of it) to design ur trading systems. Based on that u statistically arrive at signal points for the future. I understand u said that the reason for rise and fall of prices is unknown. But it is widely believed that external factors do have an effect.What i do not understand is how do u set signals based on historic data alone and not consider external factors. When u say u do not know what is causing the price to go up or down, how can external factors be ignored and we rely on the historic data alone. Or are u assuming that the special events of the past hv a similar (if not the same) effect on the prices as the new one?
First, I do not aarive at signal points for the future. I arrive at signal points for the present moment. I have no idea whether the signal will be profitable or not, simply that it stands a chance of being profitable. Statistically, I can only say with a certain degree of confidence that there is such-and-such chance of making money. Uncertainty always prevails
External events do impact stock prices, they do so all the time. What I meant was that we do not know which external factor caused price move. Generally speaking, markets, with their milions of participants, anticipate things to happen and discount those in the price. For example, last friday, in India, inflation came at a high 11.63%, higher than the expected figure of 11.44%. Markets should have taken this negatively and gone down, but it went up 2.5% instead. On other days, the same information might have caused markets to lose 3%. Why did the markets go up. Clearly not because inflation came higher than expected. It was some other reason. Had the markets gone down, people would have said it was because of inflation. Reasons for market moves exists, simply that we dont know exactly what are these reasons. But we still try to attribute reasons to market moves.
I try to build systems that are based on general principles and that do not require complex and numerous variables. The future is never like the past and the impact of unknown external events can never be built into any systems. We dont know what external event will come up, forget the impact of that on prices. But since I use simple systems that operate on general principles, and avoid too many degrees of freedom, the negative impact of external events gets adjusted in the performance through the process of losing a small sum of money without damaging much the performance parameters.
At the end of the day, any event will either have a positive, negative or neutral impact on prices. There is no fourth option. We are only bothered about the negative impact since a positive impact is welcome and a neutral impact does not matter. So I trade with risk control.
At the end of it all, if the world hs really changed, my systems will start giving performance much outside the historical performance. In quality control, one odd outcome can go out of range. But when a 6-sigma process starts behaving like a 5-sigma process, you know something is wrong. Systems also degrade and need to be repaired or replaced. Having a system that resembles market behaviour is less likely to degrade. And then I use multiple systems to avoid dependence on any one.
Post a Comment