CHAPTER 6
Quality Data
All that we have talked about so far matters very little, however, if we don’t have
the right and appropriate data to work with. The truth of the matter is that several
of the largest data suppliers have it all wrong. Look at Figure 6.1, which shows the
closing price of Microsoft as it actually was reported back in 1986, Microsoft’s
first year as a listed company. Note how it wiggles and squiggles, with hardly any
two adjacent closing prices being the same. (I downloaded this data myself from
Microsoft’s Web site.)
Now, look at Figures 6.2 and 6.3, which are taken from two different and
very popular data providers. These charts show their respective versions of how
the Microsoft stock moved during its first year traded. The charts differ from the
one in Figure 6.1 because of the added high and low prices, but two important
things must be noted here: First, because the Microsoft stock has been split sever-
al times, the price of the stock in these charts is much lower. This is all fine and
well and can be dealt with, using proper system testing and design.
More alarming, however, is the fact that most of the wiggles and squiggles
are gone: Instead, the closing price seems to jump between a few very distinct
price levels. For each split, these data vendors also adjusted the new and lower
split-adjusted price to match the nearest tradable fraction. For example, if the his-
torical stock price came out to $4.76, it was adjusted down to $4.75, because
$4.76 was not a tradable price. Over time and many stock splits, this smoothed
out the data. Note also that even these charts differ—one of the data providers
hasn’t even done the smoothing correctly. Figure 6.4 shows the data from yet
another data provider. Although this is not as bad as the other two, it still doesn’t
look perfect.67Copyright 2003 by The McGraw-Hill Companies, Inc. Click Here for Terms of Use.