Market Bottom a la Mamis

Discussion in 'Technical Analysis' started by PetaDollar, Dec 4, 2008.

  1. As far as I can tell, no sign that the market has bottomed.

    Here is the 2002 bottom.


    [​IMG]
     
  2. Now here is the current situation:

    [​IMG]
     
  3. Blue is the ratio of stocks making new highs to new lows (I believe new 52-week highs to new 52-week lows.) Red is the S&P 100 (you can look at the DJIA or whatever instead, doesn't matter).

    According to Mamis one tip off of a bottom is lower prices but less stocks making new lows (a divergence). (This is true for market tops as well-- I remember a friend of mine complaining when the market went up but his stocks didn't.)
     
  4. Good stuff. Please post again when you see rising NHNL line.
     
  5. I've backtested those kinds of things - not as easy as it sounds to make money from it as a "strategy".

    The easiest way is to take the # of highs versus lows, or versus an average and attempt to generate a buy/sell signal - usually just for the long side of a buy/sell signal.

    The other variation is to take an average of the whole market for how many stdev above or below average the Market is - this does not beat moving average or 52 week high systems trading long only.

    You can do this stuff in amibroker using the explore function or to backtest the strategy - backtest function
     
  6. Did they have any other tip-offs? It seems the NHNL line's slope might be one.

    I've been looking at the 2002 bottoms on the charts, the mad rush and exuberance over bargains was apparent after the bottom was hit, the gaps on SPY continually in the days afterward especially seemed telling.

    I think maybe the NHNL's slope is telling the same story.
     
  7. Mamis discusses and shows examples of using data (as opposed to indicators which are functions of price, such as the MACD) to identify when a trend is probably near the end. Three of his books are "The Nature of Risk", "When to Buy" and "When to Sell". Some of the stuff is outdated but much of it is applicable. There are no trading strategies. As always it is up to traders to show something tilts the probabilities in one direction and build a methodology.

    Anyway...

    ...If other folks want to post other data and market bottoms to compare to the current situation, this thread would be a great place.
     
  8. p.s. one thing to notice about the previous 2 charts: scale on the left. Currently we have been pegged below 0.10 for some time. In 2003 0.10 was just touched a few times. I can't help but look at that and think the market is in far deeper trouble this time around.
     
  9. Corey

    Corey

    That, or people FEEL we are in much bigger trouble. It is tough to say ... I certainly won't claim that I know how the world economy works.

    In my research, I have found that high levels of correlation can be an excellent indicator of intermediate highs and lows. I don't use the NHNL, but I use something fairly similar. I look at it on a sector and market level, and it has done pretty well keeping me out of buying at intermediate highs and selling at intermediate lows.

    While it is a proprietary indicator, it is pretty easy to reason out if you guys want to make your own. Basically, you just want to develop a proxy that will identify when stocks are moving in the same direction. I started off trying to use a correlation matrix, but it was a waste of time. For each sector I look at, I now just use a handful (20-30) prominent stocks and look at some simple trend indicators.

    Irrational exuberance and fear are herd behaviors. Trading with the herd is a good thing to do during normal market conditions, and is how we get 'trends'. Unfortunately, capitulation of trends typically means that the entire herd is wrong -- the market is saturated too greatly in one direction, and we get exponential moves. We want to make sure that we are either fighting the other direction, or at the very least, not joining in.

    Here are my levels as of yesterday. No miracles here, just some more data to look at. The end value is my proprietary signal, and you can basically just read it as a measure of 'exuberance' or 'fear', where 0 is fear and 1 is exuberance.

    XLB, $21.53, 0.25
    XLE $43.26, 0.02
    XLF $11.87, 0.28
    XLI, $22.12, 0.37
    XLK $14.46, 0.26
    XLP $23.05, 0.31
    XLU $28.29, 0.22
    XLV $24.36, 0.25
    XLY $20.49, 0.74

    Looking at these numbers on their own doesn't tell a whole lot ... typically, you have to look at them in conjunction with previous weeks -- but we can still glean some info.

    For example: would I short XLE right now? Not a chance in hell. On the flip side of the coin, XLY looks like it might be getting for a pull back sometime soon...especially since on the 1st it had a reading of 0.08. I certainly wouldn't go long it.

    These signals also give me a good idea as to whether any trend signals I get are due to actual new trends, or just blips. If I got a 'long trend' signal on XLY right now, I would probably look and say, "why don't I wait until it pulls back a bit. If the long trend signal still exists, then we can try going in."

    To take a closer look at how this signal might give you some insight into ebb & flow, lets look at XLE in the past month or so...

    2008-11-12, 44.6, 0.07
    2008-11-13, 49.84, 0.61
    2008-11-14, 47.77, 0.33
    2008-11-17, 46.78, 0.20
    2008-11-18, 48.09, 0.25
    2008-11-19, 45.6, 0.15
    2008-11-20, 39.96, 0.0
    2008-11-21, 44.42, 0.05
    2008-11-24, 47.63, 0.28
    2008-11-25, 48.39, 0.33
    2008-11-26, 51.1, 0.61
    2008-11-28, 50.28, 0.53
    2008-12-01, 44.93, 0.05
    2008-12-02, 46.45, 0.20
    2008-12-03, 46.48, 0.15
    2008-12-04, 43.26, 0.02

    Here, we can see how intermediate lows are definitely marked. I still need to do some research on how readings may change in bull and bear markets (for example, getting above 0.6-0.7 seems to be hard in bear markets, while below 0.3 is hard in bull markets), but you can still get a good sense of what side of the trade you probably want to be on.

    See how on 11/20, we get a reading of 0? That tells us that we might be getting close to a bottom. See the rally on 11/21? Too big? Well, our indicator barely moved, telling us that there might be a whole lot more follow-through. And follow-through she did -- all the way to $51 ... getting near typical bear-market high readings. Probably a good point for exit.

    NOTE THAT THIS DOES NOT WORK IN EXTREMES! If you don't play with leverage, you might be okay -- but just like most 'reversion' strategies, NORMALITY IS A DANGEROUS ASSUMPTION.
     
  10. Corey

    Corey

    And only to prove that caveat at the bottom of my post, XLE is down 5.76% as I am writing this...
     
    #10     Dec 5, 2008