Searching through a Scan the LPS . LPS = Last Point of Support.

viktorzito

Hello My name is Victor Rendon. I m a follower of the Wyckoff Method. Wonder if anybody has developed a scan to find the LPS that is The Last Point of Support. Any advice regarding this issue, will be greatly appreciated.

markd

Well, in real time, you don't really know if a last point of support is in fact a last point of support.

But, if you generalize an LPS as a down leg occurring within the bottoming trading range, you might be able to capture some candidates for LPS.

A strategy for scan construction might be

price is below an upper price channel of some length, and above its companion lower price channel.

the upper price channel is at its lowest level over some length longer than the channel length
or
the lower price channel is at its lowest level over some length longer than the channel length

(the above "or" statements account for the long term decline preceding the accumulation phase trading range; the channels have come down with price, but haven't started to move up yet, so they are at their lowest level in some time)

Fast Stochastics of some length (e.g. Fast K 20,1) is below 20

(this tells you short term price is in a down leg)

The trick is to find parameters for the channels that work reasonably well. There is about zero chance that you will come up with a scan that captures every stock nearing the end of accumulation because the accumulation periods vary so much in length and amplitude. So, you would need to review a lot of past examples of trades you would like to have captured, create a chart style with channels and Fast K and see what parameters capture more of them. Even a good scan will capture junk, because the same conditions will be true of stocks not in accumulation, but that's OK as long as it also captures what you want.

@lmkwin might have suggestions for P&F scans.

lmkwin

I was just talking about this the other day but talking about PS, SC, and AR and their relationship to the CSS part of my Dynamic CSS pattern studies.



So.... on LPS, it going to be something that has been showing wider based positive development with a Sign of Strength (SOS) and then pulled back.

In the PnF world, the types of patterns are going to be the multi top breakout like

and [
[PnF Triple Top Breakout is true]
or [PnF Spread Triple Top Breakout is true]
or [PnF Quadruple Top Breakout is true]
or [PnF Bullish Catapult is true]
or [PnF Bullish Triangle is true]
]

These are securities that have broken above prior multi top resistance. The LPS should be pretty obvious on these types of patterns.

I would also suggest running different periods moving the pattern back so that you can also look at how the chart is developing after the patterns are shown as well. You can do that using the calendar function on the workbench. Or you can offset the days in the scan. Example below will look for those same patterns a couple weeks ago. PnF patterns are not time based so you can push them back further to see if the subsequent action is a back up to the creek or some other Wyckoffian term.

and [
[10 days ago PnF Triple Top Breakout is true]
or [10 days ago PnF Spread Triple Top Breakout is true]
or [10 days ago PnF Quadruple Top Breakout is true]
or [10 days ago PnF Bullish Catapult is true]
or [10 days ago PnF Bullish Triangle is true]
]

Remember that PnF patterns have no time consideration and the scan engine only scans point and figure patterns based on the traditional pnf scales. Traditional scaling has a tendency to favor the higher priced issues as reversals are less of a percentage on the higher priced stocks. A 3 box reversal on a 25 dollar stock is 3. On an 80 dollar stock is 3 as well. 12% on the $25 and 4% on the $80. So the patterns are a bit "easier" to achieve on the higher priced issues. The Patterns are equally valid. They reflect the same thing. It's just something to be aware of. On the lower priced issues there is bigger movement required percent wise.

As additional filters, Wyckoff wanted stocks to be in showing positive trend so adding a price above an intermeditate MA is probably a good idea.

and [close > SMA(150,low)]

You can also limit the amount that the stock has moved since the pattern by comparing the current price and the upper price channel from a couple weeks ago. The example below says that the current close isn't more than 10% above the 45 day max high from 10 days ago

and [Close / 10 days ago Upper Price Chan(45) < 1.1]

But.... Wyckoff is better identified, in my opinion, using the time charts and filters. Like Price Channels (it would be great if you could scan the Price Channel based on Close prices also), Moving Average and Moving Average Envelopes, and the ZigZag (it would be great to be able to scan on these) indicators.

Something else to keep in mind is that you would want to run this scan against a pool that had shown the Selling Climax and Automatic Rally prior to this. So you'd need a scan that looks for the Selling Climaxes in the past several weeks/months and then run the above against that list.

Bruce Fraser did a series of blog posts back in 2015 regarding all the different aspects to consider in each Phase of Wykoffian analysis. This link is to the summary with all the different links to the other articles.

https://stockcharts.com/articles/wyckoff/2015/11/wyckoff-walk-around-the-clock.html