The following strategy below is strictly my own approach. It is not a recommendation to follow these steps, it is only a description of my own strategy. The BMW Method forms a portion of my overall investing strategy; it is not the only element. I highly recommend subscribing to the Motley Fool discussion boards and reading The BMW Method board.
My implementation of the BMW Method guides about 1/4 to 1/3 of my total portfolio at this time. In times where I have a higher expectation for US equities, it will likely be a higher fraction.
I will start buying a stock after two things happen:
If I am reasonably confident about the stock at this point, I will buy an"initial purchase", which may be 1% of my total portfolio if I am fairly bullish on equities in general, but may fall to only about 1/4% if I am not too enthusiastic about equities and would prefer to be in other asset classes (as I am now, in late 2004).
After the initial buy, the stock's price could still fall. If so, and I believe my initial confidence in the company holds, I will purchase more in up to 3 additional buys, substantially increasing each purchase. This is to maximize the final return  the more stock bought at lower prices, the higher the return when the price recovers as expected.
If the Initial Buy amount is x shares, the average basis is 100% (of course), and total investment is $I (x times initial buy price), the following purchases are:
Buy
Threshold:
% of Initial Buy Price 
Buy
Amount

Total
Shares

Average
Basis

Total
Investment

90%

2x

3x

93%

2.8
* I

80%

4x

7x

86%

6.0
* I

60%

8x

15x

72%

10.8
* I

If the stock drops by 40% from the initial purchase price and I started with a 1% initial buy, I will have invested 10.8% of my total portfolio in that stock through all four purchases. This is a lot to concentrate in one stock, but should happen very rarely, and when it does happen to a quality company, a large investment is well justified. Back testing shows this magnitude of a price drop may occur less than once a year among the entire universe of quality companies, and then typically only when the overall stock market falls significantly (2001, 2002). The biggest gains (in back testing) were on stocks that dropped the most and had all three additional buys. Clearly, confidence in the company's future prospects must be high to engage in this buying approach.
After the initial buy, most stocks will at some time drop to the 90% threshold, a few to 80%, and very very few to 60%.
Once the stock is purchased, the question is when to sell. A study of investing masters generally turns up one common thread: that the purchase price is extremely important. The BMW Method helps greatly in finding excellent purchase prices. Less consistent is when, or if, to sell.
Assuming I eventually do want to sell to collect profits, waiting for a +2 RMS event to sell is unlikely to provide the best returns because of the length of time: it may take 10, 15, or more years for the stock price to rise above +2 RMS. However, stock prices fairly regularly recover within 13 years to the Average CAGR line and frequently overshoot to +0.5 or sometimes +1 RMS within such a time frame. Therefore my sell strategy would start with following the stock until it nears the Average CAGR line. I would plan to sell a little at or near the Average CAGR line, and start selling more heavily as it moves toward the +0.5 RMS range. If the stock drops and stays below a sell range but above a buy range (based on RMS and Return Factor), I would hold.
Of course, this approach assumes the company continues to be materially in the same business and in the same condition throughout. A significant change in the company's prospects would override the statisticsbased indicators for either buying or selling.