When to buy into a "significant correction"


So now you know how the Medium-Long Term Model predicts “significant corrections” (explanation here). Here’s how the model buys into significant corrections.

  1. If the S&P falls 15% (minimum MAGNITUDE) in at least 2 months (minimum TIME), BUY now. Essentially, buy when the definition of a “big correction” is met.
  2. If the S&P falls more than 20% in less than 2 months, BUY now (regardless of TIME). This essentially overrides the previous rule.
  3. If the S&P spends more than 6 months (126 trading days) without a new high and then it makes a new bull market high, BUY now. Essentially, buy when the definition of a “long consolidation” is completed.
  4. If the “significant correction” is triggered by the liquidity indicator, BUY when the liquidity problem disappears. (The S&P hasn’t made a new high in 6 months AND the liquidity problem no longer exists (i.e. margin debt change – S&P change no longer exceeds 30%).

*The liquidity problem = When the “12 month change in NYSE margin debt” minus “12 month change in the S&P 500” exceeds 30%.
As you can see, you basically just BUY when the minimum requirements for a “significant correction” are met. Why don’t we use other indicators to helps us get closer to the exact bottom of a “significant correction”?

Because we can’t

We’ve tried countless indicators before. Bollinger Bands, % distance from moving averages, RSI, # of days, etc.
We were never able to find a consistently more accurate way to catch a significant correction’s exact bottom. Whether a “significant correction” is 15%, 17%, 20%, or 23% is completely random.
For example, using indicator X might have allowed us to catch the 2011 significant correction’s bottom more accurately, but would have resulted in me getting a worse BUY $ during the 2010 significant correction.
That’s why we said “the stock market falling -15% means that the market is already bullish from a risk:reward standpoint.” Focus on risk:reward and just BUY when it supports your side of the market. It’s not possible to consistently and accurately catch exact tops and bottoms.

8 comments add yours

  1. Troy,
    Some of these rules seem subjective. I await your spreadsheet with formula and buy/sell dates. I assume they will match your published buy/sell record.
    Thanks, Dave

  2. Hi Troy
    I was wondering when to buy if the SP500 doesn’t fall 15% or more (maybe just 10%) and neither of the “buy into a significant correction” criteria is met?
    Should I just buy if SP500 exceeds the sell point (significant correction) again?
    Magnus

    • Hi Magnus,
      I cannot advise you on your individual financial situation (regulations and all), but for the record, I am not a buyer here.

      • I maybe came of a little wrong. I meant that if your model indicates a significant correction and you sell and afterwards neither of the 4 “buy into a significant correction” criteria is met. What do you do then?

  3. Hi Troy,
    as I understand it we are in a significant correction. The model missed predicting it in this case, 3rd time this happens I believe, but the model indicates that if we weren’t already long we should be buyers at the -15% level?

Leave a Comment