Is the Bull Run Over?

3 thg 8, 2022

Throughout the years following the Great Recession, many market analysts have warned that the bull run was ending.

Here's one such article from 2016: yahoo.com/lifestyle/how-much-longer-can-this-bull-market-last-124914773.html



So naturally, an important question for traders is how to objectively detect whether or not the bull run has ended by using charts.

My chart above aims to do that by using statistical tools. The chart uses a log-linear regression channel on the monthly chart of SPY to measure whether or not the bull run is intact. To create this log-linear regression channel, I added the "Linear Regression (Log Scale)" indicator by Forza . I also added the "Linear Regression Formula" by alexgrover to better gauge smaller scale trend reversals.

I modified the log-linear regression channel settings to include the entire period of the bull run following the Great Recession. More specifically, my look back period is from the bottom of the Great Recession (Count: 162, since it occurred 162 months ago). I kept the standard deviation at 2. A standard deviation of 2 means that this channel is likely to contain 95% of all price action.



Here are the ways that I use this log-linear regression channel to determine whether or not the bull run is still intact:

• If price closes below the lower channel line, the next monthly candle must move back up and, at a minimum, tag the lower channel line. (This shows that buyers are coming in to buy the dip)
  
• The linear regression line (the thin oscillating red line) cannot fall below the lower channel line of log-linear regression channel at the time of any monthly close. (If this line falls below the lower channel line it could resist price as it attempts to re-enter the channel, which in turn could signal an end to the trend).
  
• Once price closes below the lower channel line, and then recovers to close above the lower channel line, price cannot then close below the channel again without first reaching the mean (red center line of the channel). (See below for illustration)
  

This last rule is important because it signifies that there were not enough buyers who were interested in buying the dip so as to enable price to recover to the mean.

If any of these rules fail, then it is a sign of weakness and the bull run that has been in place since the Great Recession may be ending.

My chart also shows overthrows, or periods when price thrusts above the 2nd standard deviation from the mean. (See chart below)



When the linear regression line crosses above the upper channel of the log-linear regression line that sends a signal that we are near a market top. In November 2021, the monthly SPY candle formed a bearish inverted hammer, which sent an additional signal that we are near the top. These signals are opportunities to become more defensive (e.g sell call options, begin to trail and narrow your stop losses for long position, stop adding new long positions).

Some traders may ask: What is the difference between a log-linear regression channel and just drawing a channel using candlesticks closing prices? The answer is that in the case of a log-linear regression channel the channel is derived mathematically from price action using a mean and standard deviations. One benefit of this method over drawing a channel using price action is that you can better detect overthrows and underthrows. This is to say that rather than drawing a wider channel to include all closing prices, a log-linear regression channel draws a narrower channel and signals high-probability buying and selling opportunities when price closes below or above the channel, respectively. Another difference is that a regression channel will not be as skewed by an outlier candlestick.

Aside from helping to determine if the bull run is over, this chart can help you decipher whether or not bullish predictions are realistic. For example, there are bulls on Trading View calling for SPY to reach 600 next year, unfortunately, that's unlikely to happen. In fact, based on this log-linear regression channel, the probability of that happening is less than 2.5% because it would require SPY to thrust above the upper channel line (2 standard deviations from the mean) of the log-linear regression channel. (See below for a diagram)



In summary, the stock market may seem like a roller coaster that randomly takes drastic swings, but the highs and lows of these swings are quite predictable.