Apologies for the hiatus in posting. There’s a lot of work that’s happening behind the scenes; more details later. Today we bring you an important piece in Technical Analysis: Pivot Points.
When you trade a market you sometimes have no basis for what the stock is supposed to be doing in the short term. An interesting way to look at levels on a chart is to look at points of support and resistance. We’ve seen trendlines and support lines, but there’s another interesting, short term level calculation: Pivot Points.
The Pivot of any day day depends on the previous day. The formula is:
Pivot = (Close + High + Low)/3.
The reason you don’t take the open is that the open can be quite volatile and unreliable.
In India we should actually give a higher rating to the close, because it is the weighted average of all the trades in the last 30 minutes – so the chances of a freak trade value are substantially lesser.
A better pivot point would be:
Pivot = (Close*2 + High + Low)/4
This gives a double weightage to the Close.
Now a pivot is supposed to help you understand where the supports and resistances lie. Here’s the standard method to calculate pivots:
Support 1 = Pivot – (distance of Pivot from yesterday’s high) = Pivot – (High – Pivot) = 2P-H.
Resistance 1 = Pivot + (distance of Pivot from yesterday’s low) = Pivot + (Pivot – Low) = 2P-L
(Remember, a Support is where a stock might bounce back if it falls down to. A Resistance is the point where a stock is likely to stop going up)
Subsequent levels are S2, R2, which are:
S2= (Pivot – R1S1 Range) = Pivot – (R1– S1)
R2 = Pivot + R1S1 Range = Pivot + (R1– S1)
These are usually enough, but sometimes you need the next set of support/resistances, which are essentially another range away.
S3 = Low – 2*(H-P)
R3 = High + 2*(P-L)
You have to take today’s data to calculate pivots for tomorrow. So in effect, pivots are a day old. Calculating the pivot points and the S1/R1 levels gives you an idea of the range in which the stock could operate. If it breaks through S1, will it hit S2? In effect,
Let’s do this with L&T:
(All numbers based on previous days data).
In effect, today’s pivot (11 Nov) was 1619. The High and Low of the day were 1640 and 1617 which lie around the pivot point. Support 1 was at 1593 (which the stock didn’t fall till) and Resistance 1 was at 1645 which the stock didn’t rise to. This was a slow day.
Look, though, at 7th November. The stock went to R2 (1691) since the high of the day was 1692. From there it fell all the way down to S1 (1663) and then S2 (1648) bounced from there.
This gives us an idea. If a stock goes through S1, how likely is it to hit S2? The answer seems to be in favour of S1 working as a barrier;
• For L&T, for the last two years (491 trading days) there were 197 days in which S1 was crossed
• Out of these only 63 days did we see S2 hit. (32%)
• Similarly at the other end, R1 was hit on 245 trading days
• But of those, only 101 days saw R2 being hit. (42%)
(Note: we have only Open, High, Low, and Close. We don’t know if S1 hit after the high of the day was established or not. Our assumption is to take the worst case – that S1 was breached, and the stock reversed to hit R1 later, if it’s high for the day is above R1)
The win-loss ratio is in favour of “defence” of S2 or R2 – that is, the price might cross S1 but never hit S2, which has happened 68% of the time. And price may cross R1 but never cross R2, which is 58% of the time.
One way to define a trading rule could be:
• If the stock hits S1 or R1, set up a position that expects S2 or R2 to be “defended”. The win-loss ratio is 68% / 58% on the back-test.
• This can be done through options. You can write a put near S2 (or write a call near R2) expecting it to hold. The level S2 or R2 can be the stop loss, on an intraday basis.
• It’s fiendishly difficult to produce a back-test using options, but a rough calculation through just futures shows that going short at R1 is likely to be fruitful.
This kind of rule is only intraday. It will require a lot more testing – over a larger number of years for instance. And then, it will need intraday data to be used to see if such trades were actually available or not. Finally, the back test should be done across multiple stocks, not just L&T.
Here’s another hypothesis: If a stock falls to S2 or below, it’s likely to fall further the next day. If a stock hits R2, it is likely to go up the next day.
To test this, we can run it on L&T again, with the rule that:
• If S2 has been hit, short the stock (A)
• if R2 has been hit, go long (B)
• Measure against the next day’s close (no stop loss at all)
This gives us a whopping 36% return on the short (A) above, and an even bigger 43% on the long (B) above, between November 2012 to November 2014! This is much easier to implement and trade. (At 3:15 PM see if S2 or R2 are broken, if so, set up positions for the next day)
Testing this on HDFC Bank gives us a return of 8.3% (A) and 8.5% (B). This doesn’t sound like much but it’s a trade that lasts a day – so your capital isn’t blocked for too long. (And you can amplify this using options)
Other stocks will have different back-test results, but this is just so you get an idea. many stocks like Sunpharma and ITC don’t exactly conform.
Pivots can be useful points of reference in a stock. They can be used to trade in a system, or just determine stop losses. But proper back-tests and cross verification with intraday data is key to the process.
Other ways to use pivots will be to do breadth indicators. How many stocks fell under their S2, while the index went way below S2? This may give you an idea of whether the move is “secular” or just trouble in a few stocks.
Nothing in this newsletter is financial advice and should not be construed as such. Please do not take trading decisions based solely on the matter above; if you do, it is entirely at your own risk without any liability to Capital Mind. This is educational or informational matter only, and is provided as an opinion.
Disclosure: The authors at Capital Mind have positions in the market and some of them may support or contradict the material given above, or may involve a direction derived from independent analysis.