I have no idea who came up with the notion of overbought/oversold conditions. Market makers have probably appreciated the gifts from unskilled technical traders who thought these levels could be easily traded. There is worse than that: some even think oversold/overbought conditions can signal market bottoms/tops.
Classical technical analysis has failed traders. Despite that, the backfire effect cognitive bias prevents many from admitting that. Due to the fact that the basic principles are easy to learn, many are still using this failed method. Some even offer to teach new traders how to draw lines and use simple indicators. This is ludicrous given the evidence of the failures.
In this article we consider the preposterous notion of overbought/oversold conditions. The main problem with these conditions is that the indicators used to identify them involve only price but not time. As a result and as everyone with skin-in-the-game knows, markets can stay overbought or oversold for extended periods of time and these conditions are not easily tradable. But more importantly, it takes only a couple days of subdued price action for these conditions to disappear.
An example of an indicator that is used to identify overbought/oversold conditions is the RSI, usually calculated over 14 days in the daily timeframe. However, the RSI is a momentum indicator and its use in trading as an oversold/overbought indicator leads to a contrarian strategy that can fail miserably during strong uptrends such as those formed in equities markets. Certain variants of the RSI indicator can work well for buying dips assuming markets are mean-reverting but any other dip-buying method is as effective under those conditions.
I will just present some examples in this article that illustrate my points. The motivation came from tweets yesterday pointing to the fact that the Russell 2000 ETF (IWM) RSI(14) has stayed above 80 in the last three trading days, as shown in the chart below.
The above chart shows the number of days RSI(14) has stayed above 70 in the first indicator pane and above 80 in the second indicator pane.
My question is: What is this supposed to mean? Buy or short IWM? Are three days with RSI(14) above 80 a signal of a market top or a good signal for a quick short trade?
Usually there is no clear answer from those who point to these signals. In addition, there can be no backtests because this is a sample of one such 3-day period in the history of IWM. So I am compelled to look at other securities for similar conditions and how price action has reacted forward. Below is a chart of MCD.
Last June MCD price action was overbought according to RSI(14) > 70 for 38 days in a row and for 25 days in a row according to RSI(14) > 80. Those who shorted the stock after 3 or 4 days of seeing an RSI(14) above 80 lost big time. The market just ignored those indicators like they did not exist.
The SPY ETF RSI(80) stayed above 80 for three days in a row in September 1995 as one of the longest uptrends in stock market history was ready to begin, as shown in the bottom pane of the chart below.
But why going that far back in the past? During February of this year, the DJIA RSI(14) stayed above 80 for 5 days with the index just below 21,000. There was only a short-term top but since the index has increased by about 1,500 points without any new major overbought conditions except a brief encountered in the beginning of August that could only be exploited by fast short traders but at high risk.
The message is that overbought/oversold conditions signals are ambiguous and usually followed by continuation along the primary trend rather than reversal. Some technical analysts may respond with an example or two where the conditions signaled reversal but that is usually selection bias. The problem with most technical analysts that draw lines and indicators is that they are always compelled to pick some chart where some idea worked in the past. I usually pick charts where the idea did not work in an effort to show that isolated examples do not reflect the true expectation from using these methods. My analysis indicates that the longer-term expectation from trading classical technical analysis is zero before trading friction. When transaction cost and slippage are added, the result is a loss with magnitude depending on time-averages. Some may be luckier and last longer to even make a profit but some others may be ruined in a few weeks or months depending on capitalization and risk management method used.
If you have any questions or comments, happy to connect on Twitter: @mikeharrisNY
Charting and backtesting program: Amibroker
Technical and quantitative analysis of Dow-30 stocks and 30 popular ETFs is included in our Weekly Premium Report. Market signals for longer-term traders are offered by our premium Market Signals service. Mean-reversion signals for short-term SPY traders are provided in our Mean Reversion report.