Divergence Pays Off, and some DSM Targets

The divergence between price and RSI(14) which I pointed out two days ago appears to be resolving. Price has been making weaker and weaker highs. When that happens, somethin’s gotta give, usually in the form of price falling, occasionally in the form of a huge burst upward on high volume. We may be going for the more traditional resolution:

SPY 06/20/07

I also said I was on the “cusp” of going fully to cash (wherever the hell I came up with that word), and as price opened up, dropped, retraced to yesterday’s close, then failed again, my long positions went bye-bye as the SPYders came back through 153:

SPY 06/20/07

I did not flip to short. I’m currently trying out the Dummyspots Swing Method® which I’ve developed over the past few months.

The DSM (Dummyspots Swing Method) ®

This idea started evolving when I tested the TradingMarkets R2 Method using Ameritrade StrategyDesk back in early March. (BTW, Bill is using the actual R2 Method over at his site right now- check it out and follow along.

I doodled with this and changed that, trying as best I could to avoid the common pitfall of modifying a method ad infinitum to make it backtest perfectly - a pitfall which also guarantees that it’s likely to work about 0% of the time in the future… too many conditions.

Anyway, I ended up with something I think has promise, so I’ve been watching it for a few weeks, and am gonna try it for real now. As I said, it started with testing the TradingMarkets R2 Method, and I’ve settled on using a 4-period RSI for my method.

(Shhh, that’s a big secret. I’m thinking of laying claim to the “secret” “magic” number 4. Did you know 4 is the square root of 16 which is the 0.786 Fibonacci retracement of 20.3562 which in turn is the 0.618 Fib of 32.94, which not so coincidentally will be my middle daughter’s age in the Spring of 2027!)

I do have lots of rules for stops and “stop- and- reverses”, which I felt the original R2 method didn’t address adequately, resulting in too much exposure upon initial entry, and missing moves which would have been caught by a decent re-entry procedure. I also pay attention to important things like support and resistance areas on the chart of daily bars, and to a lesser extent, volume trends during these moves.

So, what are the numbers I’m watching? (Oh, I should mention this is all based solely on the S&P 500 via the SPYders). Here goes:

  • In cash for now (even though I wanna be short)

  • Long if price reverses above 151.75 (and probably growing an ulcer from it). If this long is triggered, the initial stop will be the lower of the current day’s low (that’ll be tomorrow, 6/21) or the previous day’s low (that’s today, at 150.96)

  • Long if price falls all the way to 146.90 or below, at which point I will look to enter at the most advantageous point I can spot between there and the close. Any nice intraday hammer, volume spike, hesitation or reversal will do. The initial stop for this entry would be 145.50, or about a buck fifty below the trigger point, which I consider to be a nice “loosey goosey” swing-trade sorta stop. How’s that for scientific?

  • Otherwise, I’ll be staying in cash. No shorty. And you all know how much I’ll miss that, because I just love to get small short.


4 Comments

  1. doc40 said,

    June 21, 2007 @ 12:06 am

    Glad to see you’re back and working with StrategyDesk. I too have been playing with the R2 method and started using it last week. I made a 3.5% profit (after commissions) on 11 of 14 stocks I bought early in the week and sold Fri or this Monday. My approach uses a 3-day period.

    To buy, I looked for RSI,2,D,2 to be less than 20 and RSI,2,D,1 to be above 20, same for RSI,2,D,1 and D,0, (R2 crossing up through 20) OR RSI,2 equal or less than 5 (sometimes a stock takes a nice dive). Since R2 alone can cause the buy to appear too soon in the down trend, I did a 2 day EMA for days 1 and 0, with yesterdays (1) EMA less than today’s (0). This ensures a turn up has occurred.

    To sell, I used a similar approach with the R2 going up through 90 and a 2 day EMA with day 1 about 1.2% below day 0, or
    greater. This catches a sell very close to the top and excludes those days when the stock is rising real fast.

    There are other things you have to watch out for, such as the trend, volume, price, the relationship to the 20 day MA, etc. but I like to keep it simple so I have to eyeball the chart very carefully.

  2. LP said,

    June 21, 2007 @ 6:40 am

    And you thought about quitting. I’m glad it’s not a consuming thought.

  3. Will said,

    June 21, 2007 @ 9:39 pm

    doc40 - Veddy intevesting stuff there. Please keep us informed on how it’s going. I very much like the idea of a really fast RSI (like the 2-day) smoothed with a short moving average, and it’s what I was working on at first. However, for my purposes (only taking long entries on the SPY) the straight RSI(4) gives me virtually the same signals (the levels are a little different), and it’s easier to plot on platforms w/ varying levels of sophisitication (i.e. many of ‘em suck!). Look forward to hearing more from you!

    LP - For the time being, at least, guess I’m like the “Highwayman” in that I’ll be around, and around, and around. Good to hear from ya. And good luck with the 14-day project. Your determination is inspiring.

  4. LP said,

    June 22, 2007 @ 9:21 am

    Will,

    My 14 day project is coded named “Ground Hog’s Day”. It seems to be the same day over and over again. :)

    Good to see you back and build that yatch account. I’m looking forward to doing some deep sea fishing.

    LP

RSS feed for comments on this post