No. of Recommendations: 6

The first assumption we make in using the BMW method is that 'Average CAGR' is somewhat predictive of a stock's future CAGR. In other words, there is some expectation that future 'Average CAGR' will be similar to past 'Average CAGR', and that there will be some long-term fluctuation above and below the future Average CAGR line.

I prefer to use the longest data series available to determine Average CAGR.

The longer and tighter the data series, the more confidence I have that future Average CAGR will be close to past Average CAGR.

To baseline this idea, I assume that the future Average CAGR line will start from today's Average CAGR line (at the 0 RMS price).

However, I assume that the slope of the future Average CAGR line will be less than the Average CAGR line based on past price data. How much less slope depends on the length of the past data series (e.g. 40 years), and the amount of data scatter as measured by the spread between the RMS lines.

The way I have been doing this is to determine the slope of a line drawn from the 0 RMS point at the start of the data series - the left edge of the graph - to today's -2 RMS point.

I assume that slope and the future Average CAGR will be in the same ballpark, with the conservative assumption that the stock won't grow as fast in the long-term future as it has in its long-term past.

Sheridan's BMW charting tool had a feature to graphically determine this slope by connecting the oldest 0 RMS point and today's -2 RMS point. Unfortunately, the tool no longer works.

Here is how to determine the slope using math.

We need the following input variables:

A = Average CAGR expressed as a decimal (i.e. 12.0% --> A = 0.12)

Y = years of data used to calculate Average CAGR (40 years in this example)

R = the Return Factor associated with one unit of RMS. (This is the RF of a stock when the price is at the -1 RMS line. The value of R reflects the 'tightness' of the BMW chart; R close to 1 is tight. Larger R reflects more scatter in the price data around the Average CAGR line.)

R can be calculated as (1 RMS price) / (0 RMS price) from the Klein charts.

The output variable is F (future Average CAGR) expressed as a decimal, which can be converted to a percentage.

F = ( ( ((1+A)^Y) / (R^2) )^(1/Y) ) -1

Dividing by the expression (R^2) is what causes the line to shift the line downward by 2 units of RMS over the period Y years.

Using 2 units of RMS was an arbitrary choice on my part.

I could have used /R to shift downward just 1 unit of RMS, or /(R^3) by 3 units of RMS.

Another concept I am chewing on is to use /(R^R). That would penalize tight data just a little, and scattered data a lot more, which could be helpful for the BMW method to self-select suitable stocks. More on that later.

No. of Recommendations: 3

The concept of Average CAGR assumes that a stock at some negative RMS will eventually rise back up to the 0 RMS line and higher. The question is how long until this reversion takes place.

How long should we be willing to wait?

Some people use 3 years, some use 5 years.

For low Average CAGR stocks, one would naturally want the reversion to happen relatively soon for a worthwhile CAGR on the investment.

For high CAGR stocks, the investment could easily grow at a satisfactory CAGR even if it takes a long time to reach 0 RMS again - see MO, SBUS, TJX, ROST, etc.

On the other hand, how do we distinguish between high Average CAGR stocks that will continue at a satisfactory growth rate, and busted growth stocks that will never revert?

How long am I willing to wait for the thesis to prove out?

In my bias for DGI stocks, I tried using 1 / Yield as a proxy for duration, but this has kept me out of some really good growers that never quite got to a low enough RMS. It also gave buy signals for value traps like PBI, which clearly will never get back to its long-term 0 RMS line, and busted growers like TEVA, where the dividend was soon cut, then eliminated.

Clearly, 1 / Yield can be a false signal if management holds on too long to a dividend that can't be supported by the underlying business. The market knows this before the BMW Method does.

There are some companies that 'could' pay a healthy dividend, but choose not to - see AAPL before it started a dividend, or BRK now. 1 / Yield doesn't work for these either.

I think Enterprise Value / Free Cash Flow would be a better metric here. It can catch highly profitable slow-growth companies, and fast growers with healthy balance sheets, and helps to avoid value traps.

EV drops with stock price, but responds slowly for highly indebted companies, so that a value trap like TEVA would need to drop much more in price before getting a buy signal. TEVA at $30 was not a good buy; TEVA at $12 was a good buy.

My next challenge - how to better determine a forward-looking FCF value for the EV / FCF ratio.

More on this later.

No. of Recommendations: 2

*How long am I willing to wait for the thesis to prove out? *

This has proven to be one of the big stumbling blocks of the BMWM for me. Reversion to the mean happens because there is an attractor which is the intrinsic value of the underlying company. But that is not enough, Mr. Market has to get it and it will in time but in competition with all the other investment opportunities. If you are looking at a healthcare stock, the sector might be going in or out of favor affecting the time for the stock to revert.

What this translates into is the need for long term investing horizons and patience while at present markets are choppier than ever. The first question to ask is "How long is your attention span?"

A very long time ago IcyWolf and I had long discussions about when to buy a BMWM stock. I was in favor of averaging down while he favored waiting for the uptrend before buying. At this late date I concede that he was right! While the profits in dollars might be the same, since his holding period is shorter, his CAGR is higher. In other words, buying after the bottom is better cash management. Money parked voluntarily is OK but money stagnating in a stock is not.

Denny Schlesinger

No. of Recommendations: 1

I remember IcyWolf's discussions about this. The rationale makes sense, but I never quite understood how he decided when the bottom was in, in contrast with a dead cat bounce or a bull trap. If we could pick a confident bottom after a large drop, most other concerns would vanish.

It's been a while since I read Icy's stuff. Time for a review.

No. of Recommendations: 0

From a prior post:

"R = the Return Factor associated with one unit of RMS. (This is the RF of a stock when the price is at the -1 RMS line. The value of R reflects the 'tightness' of the BMW chart; R close to 1 is tight. Larger R reflects more scatter in the price data around the Average CAGR line.)

R can be calculated as (1 RMS price) / (0 RMS price) from the Klein charts."

The Klein charts actually do the calculation for us. The chart legend has a line such as:

-1RMS (RF=1.37)

In this example, the value of R is 1.37.

No. of Recommendations: 1

*The Klein charts actually do the calculation for us. The chart legend has a line such as:*

-1RMS (RF=1.37)

In this example, the value of R is 1.37.

That RF number assumes that the price bounces back to the mean instantaneously which it won't. I calculated a yield based on the price getting back to the mean (the red line) over time

http://bmwmethod.com/screens/calculator.php

I'm not sure if this calculator is giving the right results after I fixed some other problems with the code.

Denny Schlesinger

No. of Recommendations: 1

In addition to the BMWm chart, I use traditional fundamentals. I check 10 years of: profit margin, revenue, debt levels, # of shares outstanding, dividend history, and comparative performance within a narrowly defined "industry" grouping. "Due diligence", imo, doesn't involve any technical analysis. These value metrics reveal a lot about the health of a company or of an industry. Comparing specific industry groupings is important as is the general condition of the market (over-valued?). Take "consumer staples" for example. The revenue of some of these majors (PG, Kraft, KO, CL, TR, KMB) have flat-lined for many years now. The industry group itself has hit a wall, are losing to generics or face increased competition. Pepsi at a P/E of 33? Why? Growth is negligible. So how good a measure is "price"? Interestingly, check the CAGR of Pepsi at the 40-year bmwm chart(15), and step ahead over each 5-year increment. It's on a CAGR descent down to 8 on the 16-year chart. So, in this case, I'd be very reluctant to buy PEP at an attractive RMS or return factor. Not unless lightening strikes this over-priced stock or KO goes bust.

No. of Recommendations: 1

**Pepsi at a P/E of 33?** Why? Growth is negligible. So how good a measure is "price"? Interestingly, check the CAGR of Pepsi at the 40-year bmwm chart(15), and step ahead over each 5-year increment. It's on a CAGR descent down to 8 on the 16-year chart. So, in this case, I'd be very reluctant to buy PEP at an attractive RMS or return factor. Not unless lightening strikes this over-priced stock or KO goes bust.

The GAAP P/E is a worthless bit of trivia, specially in high tech with high R&D spending. To get a P/E ratio that makes sense like it used to 20 years ago you have to strip out the CRAP that FASB has introduced in GAAP. Two primary culprits are the expensing of R&D and the expensing of stock options both of which depress earnings raising the P/E ratio but also the mark to market of securities like warrants to buy shares which makes earnings and the P/E ratio abnormally volatile.

Saul (at Saul's board) strips out these items as part of his analysis.

BTW, KO's TTM P/E ratio is 153.08. Did Mr. Market go bonkers or is FASB/GAAP living in FantasyLand?

https://finance.yahoo.com/quote/ko/?p=ko

Denny Schlesinger

No. of Recommendations: 0

Thanks, Denny, for the Saul reference. I'm reading a post of his from a while ago, I guess his "principles" of investing, and he states:

"My feelings about PE ratios: Just out of curiosity some time ago I figured the average PE ratio of my eight biggest positions. These were rapidly growing companies but the average PE was 20. Note that that goes against the MF RB idea of picking overpriced stocks, or even ones with no earnings. An exciting company with a PE over 200 or something, may do just fine over the long haul, but I've decided "Not for me." If I can find a rapidly growing stock with a reasonable PE, why buy expensive stocks where you have to hope they'll grow into their price?"

I don't know where he goes with this later on the board. I'll read more of the "Saul" postings. My guess is that he's a growth-at-a-reasonable-price investor, focusing mostly on small caps. Is this post still representative of his thinking? http://boards.fool.com/knowledgebase-newly-revised-part-1-32...

Separately, the exceptional P/E for KO is clearly an anomaly, so I'd disregard it. It's 10-yr. average P/E, though, is 34.8.

No. of Recommendations: 0

Wow, there is activity on the BMW boards!

DT

No. of Recommendations: 1

I have been using the BMW charts for years with great success. The longer CAGR's are not always the best ones to use, IMO, as some companies have slowed dramatically. Look at PG, for instance. And I'm wondering if PSA might be falling into the same category. I like looking at the long and short term charts to see what makes sense to me. I admit, there's a lot of 'feel' in my evaluation, but it works well for me. Hats off again to Mike for simplifying this charting stuff substantially for us.

MR

No. of Recommendations: 0

MR - I think most companies have slowed down, it's mostly a factor of larger macroeconomic issues and we should continue to see the further slowdown.

Some companies especially where they can/will undercut current players, will see faster than average growth.

No. of Recommendations: 3

Hey, I've been napping on this bench for 5 years. Can you guys keep the racket down?

Dan

No. of Recommendations: 6

I sensed a disturbance in the force… I hadn’t visited this drifting death star in a millennium…

*The first assumption we make in using the BMW method is that 'Average CAGR' is somewhat predictive of a stock's future CAGR. In other words, there is some expectation that future 'Average CAGR' will be similar to past 'Average CAGR', and that there will be some long-term fluctuation above and below the future Average CAGR line.*

Past stock price growth isn’t predictive. Further, the average CAGR line on stock price charts is constantly readjusted to account for recent fluctuations in price. These charts don’t show an average CAGR that the price has swung below and above (although at first glance they may seem to) but rather a bespoke median that is tailored and altered to fit the swings in the stock price. Given this revisionist methodology how on earth can an average CAGR be predictive?

Sure there will be some fluctuations below, and above, the future average CAGR line, but you can't know where that line will be, unless you have a time machine.

*F = ( ( ((1+A)^Y) / (R^2) )^(1/Y) ) -1*

“Beware of geeks bearing formulas.” —Warren Buffett

“Once you start down the dark path, forever will it dominate your destiny, consume you it will.” —Yoda

No. of Recommendations: 0

Kelbon has a point, although I disagree with his implication that long-term CAGR charts provide zero useful information.

The BMW method is simply a model, an approximation of a real-world phenomenon.

The fact that CAGR charts adjust to new data over the long term is a feature of this model, not a bug.

Price is both a signal and an indicator of value. Unless a stock is headed straight to zero, there is always some price that is 'low enough' to represent good value.

For every example of a stock that chugs along with steady growth similar to its past (like MO), there are others that fell off the rails (like AVP).

The puzzle is to understand the limitations of this particular model, and use other types of analysis to differentiate between a good value and a value trap.

Constructive discussion is always welcome.

No. of Recommendations: 1

*Constructive discussion is always welcome.*

SubGuy, as you're attracted to charts of historical stock prices with a mean, check out the print version of *Value Line*. The mean on their price chart is a multiple of cash-flow. There's lots of other useful information in each company's array too. There's probably a library within striking distance of you (if you're in the USA) that subscribes to VL.

kelbon