In this article, I will try out MACD and see how it behaves as a TA tool in real life.
Just to recap, I mentioned in the previous article that there are three ways to read MACD Buy Sell signals :- (a) MACD crossing Signal Line; (b) MACD crossing Zero Line; or (c) Divergence.
In this article, I will test out only (a) and (b) above. Divergence is a bit complicated. Maybe we will work on it some other times.
To fully understand this article, I suggest you first read my 5 May 2016 article.
2. Signal Line Crossings
As shown in the chart above, there were 4 major crossings of Signal Line by MACD during the 6 months period (the little red dots at the bottom part of the chart).
Sell signal first showed up in mid November 2015 at 46 sen (Red Circle A). Share price subsequently dropped to as low as 38 sen by mid December 2015. As such, we can say that the Sell signal worked.
After selling at 38 sen, Buy Signal later showed up at 40 sen (Red Circle B). Three weeks later, Sell signal showed up at 44 sen (Red Circle C). Once again, the trade was successful, resulted in a gain of 10% (44 sen vs. 40 sen).
After that, it was a nightmare for the Trader. Stock price started to move sideway, resulted in MACD crossing Signal Line 7 times in quick successions (whipsawing) within a period of three months (mid January until mid April). If the Trader has followed the signals faithfully, he would have lost a lot of money paying commissions.
Finally, in mid April 2016, a genuine Buy signal emerged. Following the signal, the Trader would have bought at 50 sen (Red Circle D) and hold until now (share price of 65 sen), a gain of 30%.
3. Zero Line Crossings
As mentioned in my previous article, crossing of Signal Line is a more sensitive indicator then crossing of Zero Line by MACD. What exactly does "more sensitive" mean ? Is the higher sensitivity a good or bad thing ? Let's find out.
As shown in the chart above, there were 2 crossings of Zero Line by MACD during the 6 months period (the little red dots at the bottom part of the chart).
First Zero Line crossing happened during end November 2015 (Red Circle A), triggering a Sell signal at 40 sen. Subsequently, share price declined to low of 38 sen, and rebounced shortly after that.
I would say that the Sell signal was a bit too slow (selling almost at the bottom). As a comparison, Signal Line crossing happened earlier in mid November, generating a Sell Signal at price of 46 sen (as discussed in Section 2 above).
Second Zero Line crossing happened in 3rd week of December 2015, generating a Buy signal at 45 sen. Now come the interesting part - after that crossing, MACD did not cross Zero Line anymore. Meaning that the Trader will still be holding the shares until now, at 65 sen. This round, I would say that MACD has pulled off a successful trade.
4. Concluding Remarks
The following are the key observations from the little exercise above :-
(a) Signal Line Crossing - MACD demonstrated some successes. However, whipsawing can be a big problem, resulting in unnecessary trades. Whipsawing typically happens when stock price moves sideway.
(b) Zero Line Crossing - As was widely known, Zero Line crossing is slower in generating Buy Sell signals compared to Signal Line crossing. In most cases, this would compromise the quality of the signals, rendering them ineffective. However, in some cases, the insensitivity could act as a filter to eliminate whipsawing noises. If you make a comparison between Section 2 and 3 above, you would have noticed that a Trader that uses Signal Line Crossing would have carried out 11 transactions while a Trader that uses Zero Line crossing carries out only 2 transactions, a big saving of commissions.
So what are my conclusions ? Does MACD work ? Is Signal Line crossing more superior than Zero Line, or the other way ? Can we say that Signal Line crossing is suitable for short term Traders while Zero Line is for FA Investors ?
Sorry to disappoint you - I am unable to answer any of the above questions. In my opinion, the little exercise shown in this article is too small and too limited in scope to be of any statistical significance. The findings could be vastly different if we pick another stock as subject for analysis. It is too early to jump into any conclusion.
However, the little analysis does churn out some useful insights. It gives us a feel of how MACD behaves in real life and some of the issues associated with it. I think that is sufficient for now.