So I get this e-mail that talks about pivot points and ...

>And you never heard of them, right?
No. I mean, yes, I never heard of them, but they seemed interesting so I thought I'd investigate.
The idea is to look at the variations in today's stock prices, the Open, High, Low and Close, and construct a so-called PivotPoint price, then generate some resistance and support levels using the PP and the O, H, L and C prices.

>Resistance and support? Resistance and support levels are prices where you expect the stock to have some difficulty in crossing. If they do cross, you might expect some major stock movement as in Figure 1. Note: In addition to resistance and support levels, there are resistance and supports "trends". (See Trends and PandF and Raff regression.) With pivot points, we just talk about "levels" ... as in Figure 1. There may also be several such levels, some weak, some strong. Crossing a strong level may be the start of a major trend, UP or DOWN.

Figure 1

>And what about pivot points?
Okay. We define, at day's end, a pivot point:
[1a]       PP = α₁O + α₂H + α₃L + α₄C     where α₁+α₂+α₃+α₄=1

Then we can define resistance and support levels:
[1b]       R = β₁O + β₂H + β₃L + β₄C     where β₁+β₂+β₃+β₄=1
[1c]       S = γ₁O + γ₂H + γ₃L + γ₄C     where γ₁+γ₂+γ₃+γ₄=1

>You kidding? Am I supposed to pick all those numbers, α₁ and α₂ and ...
Why not? Or, you can just adopt what others have chosen.
For example, here's one choice:
[2a]       PP = (H + L + C)/3     Note that 1/3 + 1/3 + 1/3 = 1
[2b]       R3 = (5/3)H - (4/3)L + (2/3)C     Note that 5/3 - 4/3 + 2/3 = 1
[2c]       S3 = (-4/3)H + (5/3)L + (2/3)C     Note that -4/3 + 5/3 + 2/3 = 1

>Where's O?
The Open? It ain't there ... for this particular choice.

>Example? I'm glad you asked. See Figure 2? It's GE, starting on March 10, 2005. The High, Low, Close, Pivot Point, Resistance and Support are shown ... for March 10, using magic formulas [2]. >So what do I do now? I guess you wait and see if one of those support levels gets crossed and ... >Crossed violently? Okay, a violent crossing. Then you might consider buying or selling. Of course, you may have several support and/or resistance levels so you watch to see which are crossed. Maybe you have a "weak" level which will attract your attention ... then you watch to see if a "stronger" support gets crossed later. >Weak? Strong? Are they technical terms? Yeah ... just like violent.

Figure 2

>What's the reason for calling the levels R3 and S3. What happened to R1 and R2 and ...?
Patience. I said you can have several levels. Remember?
>Yeah. Weak, strong and stronger. I assume there's a spreadsheet to do all this?
Eventually. In the meantime you can play with the spreadsheet that James sent me.
It looks like this:

If you'd like to see what happened in the past, you can play with this guy:

Note that, in early June we crossed the "weak" support level S1. We pay close attention.
Then, on about June 22 there was a dramatic crossing of the stronger level, S3.
>So we sell?
That's up to you. Remember we're looking at the past ... and hindsight is 20/20.
Here are a few more (using the same magic formulas as above):

Here are some pivot point formulas (where particular values are chosen for the parameters α, β, γ and δ R1 = 2PP - L S1 = 2PP- H R2 = PP + (H - L) S2 = PP - (H - L) R3 = H + 2(PP - L) S3 = L - 2(H - PP) The spreadsheet uses these as well as: PP = (H + L + C)/3 Although none use the Open, the calculation of the Pivot Point might use it. For example: PP = O + (1/4)(H + L + C)

Figure 3

If the formulas seem strange, we can rewrite them so they give deviations from previous levels, as a percentage of the PP, like so
(using the notation in Figure 3).

R1 = PP (1 + b%)	S1 = PP (1 - a%)
R2 = R1 (1 + a%)	S2 = S1 (1 - b%)
R3 = R2 (1 + b%)	S3 = S2 (1 - a%)

>Huh?
In place of a = H - PP, we consider a% = a/PP ... as a percentage of PP. And instead of b = PP - L, we use b% = b/PP etc..
Then it's easy to see how we can generate a jillion levels.

Of course, in addition to using today's prices, you might use yesterday's as well and maybe volume weighting or ..
>What's the best?
I have no idea. As you might expect, there are variations on variations. See, for example, Investopedia.

>I assume these pivot points guys work pretty well, eh?
I have no idea, but here's an example where, each time the closing price crosses a resistance or support level, we calculate new ones.
>New ones? Which ones, since there are variations on variations and ...?
I use the magic formulas given above for R3 and S3 and PP = (H+L+C)/3.