Students of Roger Montgomery's (Buffett's) intrinsic valuation method

[craft]

I don't understand this. The formula assumes the RoE continues forever not 11 years.

Two parts

The paid out portion is calculated as a perpetuity.

The growth part has a time frame. Because the calculation raises (ROE/Required return) to the power of 1.8 it gives a varying time frame. For example an input of 30% ROE and a 10% required return is implied to last unchanged for 11.84 years. Whilst an input of say 15% ROE and 12% required return has an implied time frame of 15.2 Years.

To my way of thinking using a formula that implies arbitrarily the growth period, which is one of the most important assumptions is really stupid. It’s even more stupid if you don’t know what time frame the formula is implying and RM has tried his best to keep that secret – lest everybody would work out that his secret herbs and spices are just crap.

 

[McLovin]

Two parts

The paid out portion is calculated as a perpetuity.

The growth part has a time frame. Because the calculation raises (ROE/Required return) to the power of 1.8 it gives a varying time frame. For example an input of 30% ROE and a 10% required return is implied to last unchanged for 11.84 years. Whilst an input of say 15% ROE and 12% required return has an implied time frame of 15.2 Years.

Thanks craft.

I'm a little unsure on how you arrive at the implied time period. Can you work through one example?

Cheers

 

[craft]

Thanks craft.

I'm a little unsure on how you arrive at the implied time period. Can you work through one example?

Cheers

Taking the 30% ROE and the 10% required Return.

Divide 30%/10% = 3

Raise 3^1.8 = 7.225 (this is the amount he multiplies equity by for the retained % amount so that he can get a value for growth. [Lots of other problems with this assumption as well]

(1+30%)/(1+10%)-1 = 18.18% . This is the resultant % of compounding at 30% and discounting at 10%

NPER calculation with PV = -1, FV = 7.225 and rate 18.18% = 11.84 Years.

So only if the company happens to only grow via retained earnings and retain the exactly same amount of earnings and apply them at exactly the same historical ROE for exactly 11.84 years and then suddenly go ex growth overnight will the formula accurately estimate the growth value. Its utter Bull.....

 

[McLovin]

Taking the 30% ROE and the 10% required Return.

Divide 30%/10% = 3

Raise 3^1.8 = 7.225 (this is the amount he multiplies equity by for the retained % amount so that he can get a value for growth. [Lots of other problems with this assumption as well]

(1+30%)/(1+10%)-1 = 18.18% . This is the resultant % of compounding at 30% and discounting at 10%

NPER calculation with PV = -1, FV = 7.225 and rate 18.18% = 11.84 Years.

So only if the company happens to only grow via retained earnings and retain the exactly same amount of earnings and apply them at exactly the same historical ROE for exactly 11.84 years and then suddenly go ex growth overnight will the formula accurately estimate the growth value. Its utter Bull.....

Thanks again craft, makes perfect sense now. What's bizarre is that the formula assumes a higher risk company has better earnings visibility than a lower risk one.

 

[craft]

Thanks again craft, makes perfect sense now. What's bizarre is that the formula assumes a higher risk company has better earnings visibility than a lower risk one.

Many things are bizarre with the formula.

And as for the name of this thread. I think you could be pretty certain that Buffett does not use any such formula. His point of view is that a company is worth the PV of its future cash flow which indicates he is using DCF.

It just a case of a Local Districts second grade lower order batsman who has a record of underperforming the index trying to leverage his name to the Bradman of value investing.

 

[New Stratos]

It's not that "Roger's Formula" assumes infinity or 11 years, it's just that after a certain point future years tend to have a tiny effect on the result.

 

[craft]

It's not that "Roger's Formula" assumes infinity or 11 years, it's just that after a certain point future years tend to have a tiny effect on the result.

Ummm.............

 

[New Stratos]

Ummm.............

My comment was actually a reply to McLovin, but go ahead, set me straight.

 

[craft]

My comment was actually a reply to McLovin, but go ahead, set me straight.

Read the other posts in response to McLovin. Do you understand it?

 

[New Stratos]

Read the other posts in response to McLovin. Do you understand it?

yes. So what is wrong with how I summarised it? If you'd like a slightly longer version:

It's not that "Roger's Formula" assumes infinity or 11 years, it's just that after a certain point future years tend to have a tiny effect on the result and the point where this occurs is nearer in time with higher ROE's.

 

[kermit345]

Essentially what your saying New Stratos is that high ROE's are not sustainable and so the higher they are the shorter period of time that it should apply to the equation.

I don't use Roger's formula so i'm not very familiar with it but this does make sense if it is what Roger is trying to achieve. Although just because a company has a high ROE and its 'unlikely' to sustain it for a number of years, i don't see why this should hurt the valuation.

Look at a company like PTM which i think is high quality. It will be punished by Roger's forumla due to the mechanics of their balance sheet and the nature of their business.

I believe a valuation equation should be almost solely based on the current earnings/debt/returns of the company as they are 100% correct and tangible. You then apply the same equation to previous or future years (estimates) and that helps you to identify a trend in the companies valuation to give you an overall picture of where their earnings/dividends are heading and hence eventually where their share price should move towards.

Thats my over-simplied take on intrinsic valuation anyway.

 

[New Stratos]

Essentially what your saying New Stratos is that high ROE's are not sustainable and so the higher they are the shorter period of time that it should apply to the equation.

Actually, I wasn't saying that Kermit It may be true, or not, but it's not my point. I was essentially just saying that when you discount the cash flow at a higher rate, the point where future years are insignificant is sooner.

I believe a valuation equation should be almost solely based on the current earnings/debt/returns of the company as they are 100% correct and tangible.

I get your point, but I think it's a bit strong to say that the current figures are 100% correct - trickery and mistakes aside, they are still only estimates of the true state of affairs - hopefully good estimates ...

You then apply the same equation to previous or future years (estimates) and that helps you to identify a trend in the companies valuation to give you an overall picture of where their earnings/dividends are heading and hence eventually where their share price should move towards.

Thats my over-simplied take on intrinsic valuation anyway.

Makes sense to me

 

[odds-on]

Long time lurker here.

I personally enjoyed value.able, not great but not bad either. To be honest all I really want in an investment book these days is lots of examples of both good and bad investments by a respected value investor. I would quite happily pay $500 for a book by Joel Greenblatt detailing his 50 best and 50 worst investments to date.

 

[craft]

yes. So what is wrong with how I summarised it? If you'd like a slightly longer version:

It's not that "Roger's Formula" assumes infinity or 11 years, it's just that after a certain point future years tend to have a tiny effect on the result and the point where this occurs is nearer in time with higher ROE's.

You discount using the required return not the ROE. The higher the RR the quicker future years become insignificant. You are compounding at the ROE rate the higher it is the more impact on future years.

 

[craft]

Roger Montgomery's crazy intrinsic valuation method

Essentially what your saying New Stratos is that high ROE's are not sustainable and so the higher they are the shorter period of time that it should apply to the equation.

You are correct, the higher the ROE the shorter implied growth period produced by the formula. You could argue that this is a good thing because higher ROE’s are harder to sustain. But I maintain allowing a formula to imply the sustainable growth period is crazy. Your example PTM is a low capital intensity business so high ROE's are potentially sustainable. Allowing a formula to dictate that it will have a shorter growth period then a high capital intensive business with a lower ROE is crazy.

The businesses where the realistic assumptions about their future most mismatch the assumptions embedded in the formula get the most mis-valued. Resulting in some great business being passed up at cheap prices and others being brought as ‘bargains’ at ridiculously high prices.

 

[kermit345]

No arguements from me craft. I'm a fan of Roger's process which is essentially Buffet's process re-worded (low debt, high returns, good management, good cashflow). But i'm not a fan of the pumping of his newly revealed skaffold or how he dangles a carrot infront of everyone about his gold investment but waits weeks to say what it is. I understand he has investors to worry about and he was probably waiting to finish his buying but to me your either all in or all out with your disclosure, not half half when you feel like it.

Totally agree with the PTM example. An equation shouldn't have the ability to discard a company that essentially fits the criteria your going after anyway. PTM has continual cashflow from FUM, high ROE, low/zero debt and good management yet the equation would dictate that due to its high ROE, it can only be sustained for a very short number of years which, if you know the company, you know is not true.

Yes PTM is probably one of the more extreme examples but it does show you how equations and formula's which you design to achieve 1 thing, can ultimately miss those goals without further investigation. Thats why I like to have my equation based on earnings and dividends while allowing a small premium/discount based on ROE and D/E. I also attribute a premium/discount based on management qualities and that essentially gives my IV which still only completes a piece of the puzzle.

 

[Macros]

@ Craft

What you are saying is reasonable, however I think you are missing the point. Two people using a similar approach which mostly come up with different figures. The exact figure doesn't matter so much; what matters is the quality of assumptions and the conclusion. Using DCF to the Nth degree does not increase certainty as it is still based on the underlying assumptions.

In my opinion the application of a process is what counts. Being able to apply a consistent (ideally successful) process in all market conditions is the key.

From the way I look at investing, it is the market expectation of future earnings that generally will drive the price at move it towards the range of a reasonable value. Being able to understand and correctly predict where earnings are likely to head is much more important than getting valuation exactly right. Therefore, factors such as the macro environment and industry trends are very important drivers of company value in both the short and long run, as they generally impact on earnings.

With regards to Montgomery criticisms, I maintain that a lack of understanding of economic factors and change remains the biggest problem. This results in his A1 system having no meaning in my view.

@ Kermit

Again, harping on economic factors, PTM is a terrible example. Montgomery, from memory, had a value of $5 or so and listed as an A1. Not sure what he has it at now. In my view, PTM has had a value of around $2.50 to $3.00 for the past few years. From the price action, it looks like it could well head to $3 - and for good reason. PTM faces massive hurdles with significant market volatility, reduced investor interest in markets and a generally terrible environment. PTM would probably be a good play when these factors reverse (could be years), but how will they go until then?

 

[Tysonboss1]

No arguements from me craft. I'm a fan of Roger's process which is essentially Buffet's process re-worded (low debt, high returns, good management, good cashflow). But i'm not a fan of the pumping of his newly revealed skaffold or how he dangles a carrot infront of everyone about his gold investment but waits weeks to say what it is. I understand he has investors to worry about and he was probably waiting to finish his buying but to me your either all in or all out with your disclosure, not half half when you feel like it.

Buffett never talks about what he is buying until the deal is complete or until he must legally disclose a holding.

 

[craft]

Roger Montgomery's Crazy intrinsic valuation method

@ Craft

What you are saying is reasonable, however I think you are missing the point. Two people using a similar approach which mostly come up with different figures. The exact figure doesn't matter so much; what matters is the quality of assumptions and the conclusion. Using DCF to the Nth degree does not increase certainty as it is still based on the underlying assumptions.

Hi Macros
I have no problem with two people reaching different valuations for a business because they make different assumptions about the future. What I have a problem with is people thinking they are valuing a business using a rigid formula.

I agree that what matters is the quality of the assumptions and that is why I think it is crazy to let the formula make the assumptions for you. I agree that DCF is only as good as your assumptions and that is where the skill is required. But if you try to avoid having to get the assumptions right by handing valuation over to a rigid formula like RM advocates then:


If the actual growth period is different to the implied growth period the valuation estimate will be wrong.

If current ROE is impacted by historical write downs or accounting goodwill the estimate will be wrong.

If the business is subject to revenue cycles the estimate will be wrong.

If the business margins are increasing or decreasing the estimate will be wrong.

If the company uses funding other than retained earnings the valuation will be wrong.

If the companies business model can utilise customer cash flow to fund its growth, estimates will be wrong.

Blah Blah Blah.......

Nobody knows the future and your assumptions may turn out to be wrong this is the reason you need a Margin of Safety when using DCF valuations. The fact that the imbedded assumptions in a formula may bear no resemblance to what can be reasonably expected for a business means you are not really valuing a business with a formula you are just generating a figure and any MOS you think you are getting by buying at a discount to that figure could very easily be illusionary. Much like a stopped watch - the formula might be right by coincidence on the odd occasion but you wouldn't want to put any faith in it if being on time is important, and if you are betting that you can value a business better then the market then your valuation is critical.

 

[craft]

Re: Roger Montgomery's Crazy intrinsic valuation method

Much like a stopped watch - the formula might be right by coincidence on the odd occasion but you wouldn't want to put any faith in it if being on time is important, and if you are betting that you can value a business better then the market then your valuation is critical.

If you are thinking about wasting your money on his stock valuation service (how much does it cost?) I would have a think about this first.

RM had MCE valued at $11.13 for 2011 as recent as May. The watch had stopped a long way from reality in that case. On the other extreme have a good think about how the formula values something like NVT.