FTSE Probabilities

The structure below represents the probability of successive up (green) and down (red) moves in the stock market on successive days. [This is worth redoing on 2 day, 3 day trading windows]

The underlying belief is that Up/Down data is what is critical for a trader in a stock. This determines the buy/sell behaviour. The size of the moves on the other hand is only critical for the trader who is choosing a stock. Those stocks with the biggest moves might attract the attention of high-risk investors looking for big gains etc.

The daily FTSE close price since 1977 is linearly skewed so the data set ends on the same price it started. Reason for this is to try and reveal more of the intrinsic up/down data away from the background of economic growth and inflation.

Chains of particular sequences were then counted through the data set. There were 6715 days in the data. This meant that for strings of 4 days (e.g. uddu) of which there are 16 possibilities the total count of all 16 possibilities was 6715. This was true of all sequences. The probability of a particular sequence was easy to calculate as: Occurrence/6715. Dividing each probability by its parent (so that udd is the parent of uddu) gave the probability that the market went up at that point.

The obvious point is to see if there are daily trends in the stock-market that drive patterns of movement. While not tested yet these data are very likely to be significant so that the distortions to the tree below represent actual features of the FTSE.

Because this is FTSE data no single group of traders is responding to the FTSE moves. This is not speculative strategy, but rather patterns in the whole market together. The tree ought to be more regular than it is!

Red means a move down in the price, Green is a move up. The higher the probability the longer the bar and the heavier so it tilts in that direction. For example, the first branch between a stock going up and down from a random entry point is biased in the up direction, as is the second move.