Skip to main content

Real traders are guided only by facts, and they ignore rumors and opinions.

For instance, our experts at moomoc work with historic price data in order to develop a good set of rules for buying and selling stocks.

In the course of this, traders examine the performance of these assets and test whether on this basis it is possible to derive a statistical advantage to then help generate profits. So where do these data come from, which data should be used, and why and how can the quality of these data be checked?

When traders are starting out, they should concentrate solely on the so-called end-of-day data, which is based on daily closing prices. This includes the following information for each stock symbol: opening price; daily high; daily low; and closing price.

Additional information is supplied with the EOD data, for example, daily trading volumes, though volume isn't a relevant decision criterion. Therefore, based on EOD data, such strategies always buy or sell stocks at the open or at the close of the respective trading day.

But now, where can one get good intra-day data to, for example, develop a trading model based on an hourly chart or a 15-minute chart? Of course, there are sources where one can obtain such time series, but there a lot of problems.

First, suppliers don't provide this information about stocks free of charge. Furthermore, the quantity of data will increase sharply.

Although with EOD data, one trading day consists of one bar or candle in an hourly chart, a U.S. trading day has seven candles. In a 15-minute chart, it is 26 bars and in a minute chart, 390.

So, a trader has to manage 390 times the quantity of data, and with a stock universe of a few thousand units over 20 or more years, it all adds up to a nice sum.

However, the main problem isn't the quantity but rather the quality of the data.

Scroll to Continue

TheStreet Recommends

EOD data are regularly maintained and verified. Providers quickly spot serious data errors, such as forgotten splits or simple spikes, and can correct these.

But how can minute bars be verified and by whom? 

With intra-day systems, traders always have to live with the uncertainty that their systems are possibly based on wrong information. That is why private traders should concentrate for as long as possible on the development of models that are based on EOD data.

Traders should also remember that when they buy intra-day data, ultimately, they are always buying a pig in a poke because they can only compare the prices of the individual providers but not the quality of the data. Nevertheless, traders can usually assume one thing: the higher the cost, the better the data quality.

At any rate, high-quality intra-day data are significantly more difficult to obtain than good quality EOD data.

Naturally, there are also computer-based possibilities for editing out the spikes in these intra-day data series. But we are talking about simple trading models, and these options are anything but simple.

Traders shouldn't allow themselves to fall victim to the fallacy that EOD systems must be bad and that intra-day models are the holy grail.

Traders might perhaps be able to improve their performance slightly with some smart intra-day entries and exits. However, if they use good EOD systems, they will generate enough performance this way.

This is because most traders forget one thing: They won't need even more performance if they have good systems, but what they do need is less risk so that they can withstand the draw-downs. 

EOD data allows traders to develop very good strategies. Intra-day systems are of course tempting, but during the first years as a trader it would be better to do without them for the aforementioned reasons.

This article is commentary by an independent contributor.