The use of technical analysis tools is normally confined to asset price data, however, there are also some cases where the same methods can be used to analyse other data types such as econometric results – normally the exclusive domain of fundamentalists.
Economists already apply moving averages, which are a technical tool, to employment and other types of data, and there are also some technical analysts who argue techniques such as Elliot Waves and Fibonacci can be applied to a wide variety of phenomenon since they represent an underlying cosmic pattern of growth and decay.
In this article I have decided to try to represent Non-Farm Payrolls (NFPs) results in a candlestick chart format to see whether it formed classical candlestick patterns. Candlesticks are a leading indicator so I was hoping they might help forecast future payrolls. The experiment yielded some interesting insights and observations.
A candlestick is made of an open, close, high and low price over a given period of time. NFPs are released once a month so have a very low frequency. The first problem was to work out how to construct the candlesticks using such low frequency data. In the end I decided to use a period of four months of payrolls so I could use the first and last month’s value as ‘open’ and ‘close’ and then the highest and lowest of the four values for ‘high’ and ‘low’.
In this it was possible to shape Non-Farm Payrolls into the ‘open’,’high’. ‘low’, ‘close’ format required to construct candlesticks on excel. Essentially to get round the problem I had to create 4-month candlesticks.
The first thing to notice is that the resulting chart displays a striking uniformity in the sequences of up and down candlesticks. On the left hand side is a long uninterrupted trend of up-candles from between May 1991 and September 2000. It is striking that within that sequence there is not a single down candle.
This is followed by a top in 2000/1 and then a decline which is represented by an equally uniform and uninterrupted trend of black down-candles between 2000 and 2003.
This in turn is followed by another uninterrupted trend of positive candles from 2003 to 2007.
Which then tops and results in another decline, represented by a line of uninterrupted black down-candles until 2010. Since then Non-Farm Payrolls has enjoyed another up-trend – again without the blemish of a single negative candle.
Another characteristic to note, is that during up-trends candles are generally quite long, and it is only during the topping out and down-turns in payrolls that the candlesticks sometimes become short and stubby, showing marginal results, either positive or negative.
Specific 12-month pattern at tops
After analysing the candlesticks during the two recessions in 2000/1 and 2007/8, I noticed that both exhibited a similar ‘topping pattern’. The pattern consisted of three consecutive candles at the end of up-trends. The first candlestick showed a much smaller-than-usual rise, the next a slightly higher but still below average rise and then the third and final candle a decline.
This pattern preceded both recessions. In the 2000 recession the first period began in May 2000 with a very small rise of only 117k which was well below average, the second showed a slightly lower than average gain of 352k and the third and final candle was black and showed a minus figure of -236k, which was the first of a sequence of declining periods, during the recession.
In 2007/8 a very similar pattern occurred, with a very noticeable slowdown already happening in May 2007 when NFP’s only rose by 20k in the whole of the preceding 4-month period. The next period showed a slightly more robust but nevertheless still below average rise of 297k in September and the final period in January showed an actual fall, in keeping with the pattern’s guidelines.
Overall the pattern describes a 12-month period which seems to mark the beginning of recessions, because after this pattern payrolls normally start a protracted decline.
The first sign payrolls were in trouble occurred with the initial squat candle showing only 117k rise in 2000 and 20k rise in 2007. These are the two lowest rises in the whole sample, and are way below the usual range of between 300 and 900.
Potential use as indicator
The NFP 4-month candlestick chart is useful as a method for alerting investors to changes in the long-term trend of the stock market. On the chart of the SP-500 below which covers the same period as the NFP candlestick chart I created note how the two major peaks in stocks coincided with the two peaks in the Non-Farm Payrolls Candle chart. Obviously this is hardly surprising given the extraordinary importance of employment on the health of the economy and therefore of the stock market.
I have also transposed the periods for the topping pattern from the NFP chart onto the stock chart using some vertical lines describing the start and finish of the 12-month topping period. As can be seen they coincide with tops in the stock market:
In both cases the NFP topping pattern picks the stock market tops almost perfectly, particularly in the 2007/8 recession. The topping pattern did not occur anywhere else in the data and – notwithstanding the fact I have only shown two examples of the pattern – could provide investors with a useful additional indicator for warning of recessionary tops.
Aid to Investors
The NFP candlestick chart may have other possible trading applications.
The first is a simple rule for distinguishing the end of bear markets, which are distinguished by the presence of two consecutive positive candles.
The next application is to use NFP candlesticks as a confirmation or non-confirmation tool. In this method the investor would use NFPs to distinguish between mere corrections in the stock market and the beginning of more protracted bear market declines.
If the stock market has corrected back a substantial degree the investor is left with the question as to whether this is just a correction of the prevailing up-trend, or the start of a bear market.
Corrections like this are quite frequent, and happened in May 1997, Sep 1997, Sep 1998, Sep 1999, Sep 2004, May 2005, Sep 2005, May 2006, Jan 2007, Sep 2010, Sep 2011, May 2012, Sep 2012, Jan 2013 or Sep 2014.
To aid the investor in deciding whether the pull-back is just a correction and stocks are likely to go higher again, NFP’s can be used as a confirmation tool.
If the pull-back is accompanied by a positive 4-month NFP candle it would probably indicate the pull-back was just a temporary correction and that the stock market would probably eventually resume its up-trend. In a way this makes perfect common sense since healthy levels of employment are a corollary to a healthy economy, which is good for stocks.
If on the other hand the pull-back is accompanied by especially weak or negative payrolls, or the topping pattern described above, then that could probably indicate a more extending bear market might be starting.
It the former is the case the investor might decide to increase their holdings of stocks during the pull-back, taking advantage of more attractive prices, if the latter they might decide to sell, or go short stocks.