Avenue Supermart: a compounding machine?

(edwardlobo) #624

Use the growth of 13pc in your spreadsheet and it will arrive at valuation today
But Dmart grow much more faster than that
They would want to have a first mover advantage
And the new debenture issue is pointing towards that

(Dinesh Sairam) #625

I will most definitely Value Avenue Supermart using my spreadsheet in the next week or a little later. So, no worries there.

I’m more interested in looking at how you Value Avenue Supermart, since you claim to have an automated excel which does the job for you.

We seem to have different approaches. So no harm in comparing notes.

(edwardlobo) #628

Yes they also research many many companies
If you are not good at maths you atleast need to be very good at populating spreadsheets
It should take one second to populate and arrive at a basic value
Will you buy everything that comes up as a value, no
You will look further
You are selling dcf as an all important decision making tool
Either be good at maths or be good and populating excel really fast
If munger would be looking at Dmart, he would have already spent his first 2 seconds to arrive at dcf and then spend his next days researching further

(Dinesh Sairam) #629

When? Please make a quote of one of my posts where I did that.

Valuation is the final frontier. You don’t Value a company and then research it. You research it, build a sense of the business and then Value it.

I sold a DCF (Or Valuation in general) as an important part of investment decision making. It’s not everything in investing, but one of the most important things.

After doing that, I simply asked for your Valuation for DMart, since I wanted to compare notes. You blatantly denied this, even though, according to you, all it takes is a click on Screener. And you accused me of finding faults in your Valuation (When I haven’t even looked at yours). I don’t Value a company unless I research and understand the business (Which takes time). Since you’re already interested in the company, I thought you already did the due research. But it looks like you haven’t. My bad for asking.

So sure, it doesn’t make sense to continue this conversation anymore. I’d still keep my promise of Valuing DMart, even if you don’t keep yours.

(atul1082) #630

We definitely get a lot of insight by healthy discussions.pl keep the good work on and do share your research work and spread sheets to benefit us.

(phreak) #631

Happened to be reading a book and came across this section about convergent and divergent problems. Thought it was relevant to the recent discussion here which was similar to finding the right answer to a divergent problem.

(ranjan rakshit) #632

(thecroc) #634

While the company can increase discounts leading to bigger volumes , it can turn internal levers to generate better net margin. I would like to see if this can trend for a few more quarters/years like this.

(Dinesh Sairam) #635

My valuation of D-Mart, as promised.

I made a blog post explaining my assumptions a little more. However, I cannot share the link owing to moderating restrictions. Subscribers to my blog can view my justifications there.

(initin) #636

Great effort.

I have just one question did you triangulate your sales number with the number of additional stores that will be required to achieve those sales ?

(edwardlobo) #637

The model uses so many forumulas, even Ramanujan Srinivasa wont be able to calculate it in his mind, let alone Charlie Munger

There is something fundamentally wrong
The total invested capital in cell I67 in year 20 is roughly 720,000 crore

Just a present value of 720,000 discounted using a risk free rate comes to the present day amount of 150,523, which is 50% over the current market cap

This does not account for any cash flow in the interim period

So the spreadsheet basically says in year 20 the liquidation value of dmart will be less than the invested capital

(Dinesh Sairam) #638

That’s a great question. And it’s important that you ask that question. In my model, Growth and Capex are linked. So, it holds that I have to justify at least one of these figures (I personally find Growth easier to justify than Capex). Here, I recall watching a lecture by Prof. Damodaran, where he addressed the question of justification of assumptions. One of his students had valued Netflix, but he felt like something was wrong, so he consulted the Professor. The Professor was of course quick to point out that he had simply projected Growth out for many years, without justifying the figures, so much so that his assumption meant that in 15 years, almost the entire world would be subscribers to Netflix. It’s not even funny how dangerous projections can become if you do not justify them.

Coming back to answering your question, yes I did, in my head. But since you asked, let me demonstrate with actual figures from D-Mart’s financials vis-a-vis my own projections. Here’s how I went about doing this:

  1. I calculated the ‘Invested Capital’ of D-Mart for the last 10 years. I define Invested Capital as ‘Equity + Debt - Cash’. D-Mart has historically used the buy-and-operate model, so we know that a majority of the capital they invested went to acquiring newer land/buildings.

  2. According to D-Mart’s CEO, D-Mart grew at 10-stores-per-year in the last decade or so. There are no exact figures given anywhere, so I have assumed that D-Mart grew at around 8 stores per year in 2009 and gradually reached 12 stores per year by 2018.

  3. Dividing #1 by #2, we get Average Capital required per Store. I understand that this is an over-simplification. “Stores” can range from tiny to the huge ones like the one in Bangalore or Mumbai, but we have no other way to get any more accurate.

  4. Based on #3, we can calculate the growth in the land/building acquisition cost. This could happen due to a number of factors, the primary being inflation in land/building prices.

  5. Now, I can fill up the figures for ‘Invested Capital’ (#1) from 2019-2028 from my own model.

  6. Working backwards, I can use the average percentage of #4 to project #3 from 2019-2018.

  7. Dividing #1 by #3, we will get Store Count growth per year (#2) from 2019-2018.

Here it is:

Again, according to the same article, DMart’s CEO plans to boost the store growth per year from 10ish to 15-20ish for the next decade. We can see both of those opinions roughly correlate in the ‘Average Store Count Growth per Decade’ column.

You can download the excel and see for yourself: DMart Growth and Capex.xlsx (11.7 KB). Numbers were taken from Screener and my model.

For the remainder of 10 years, it is simply a question of enacting the universal truth that as companies become huge, their avenues for growth decreases and hence, their revenue growth rate itself. Eventually, that growth slows down to half of the long-term Risk-free Rate in India (The reason for which I attempted to explain a few posts back).

(Dinesh Sairam) #639

I’m sorry you feel that way. If you are genuinely interested in understanding the model, visit my thread on the same. I am always open to constructive feedback. If you’re simply trying to throw stones, I’m afraid I cannot be of any help to you.

My offer still stands, for comparing my Valuation to yours. As already mentioned, I do not care about the method employed. If you have some kind of logic in justifying D-Mart’s price, I’m most interested in listening.

Nice attempt, but you are mistaken.

Market Cap is the Present Value of future free cash flows discounted at the Cost of Capital. And so, it is only logical that for comparison purposes, Capex should also be discounted at the Cost of Capital.

Rs. 719479 Crores discounted at 14.23% for 21 years becomes Rs. 43,992 Crores, which is roughly 43% of the current Market Cap of Rs. 1,00,677 Crores. The rest, as you have already figured out, will be the PV of cash flows, both positive and negative.

(Bheeshma Sanghani) #640

Hi @edwardlobo

The gap between the cmp and the actual unseen intrinsic value of a co is made up of two components. The actual gap itself + the error of investors in estimating that gap.

While errors are inevitable as the future is uncertain they cant be huge as otherwise it defeats the purpose of valuation even if you use the most appropriate tool.

If lets say 100 intelligent analysts value dmart independently and state its true value and if one was to create a normal distribution out of that and if the standard error of that distribution is unacceptable, then one can conclude that something is wrong with the estimate. The range of error has to be small - it cant be i would value dmart somewhere between 1000- 2000. Max error should be +/- 10% of your estimate.

(edwardlobo) #641

Dinesh’ value is off by roughly 25pc
What you are advocating is efficient market theory
Read up margin of safety by Seth klarman, there are many examples there of stocks trading at 1c and going on to trade in Full dollar value and sometimes double digit dollars
That book sells for more than a thousand dollars but there are copies on the net

Dinesh, I don’t want to make this into a who is correct thread
I think we have discussed enough as you would see from my number of previous posts I don’t have enough time
I don’t use complex spreadsheets anyway thanks for your spreadsheet

There are many good companies, if you feel this is expensive I would spend my time on the scripts I find cheap but please don’t be offended. If you feel you can contribute here please do so even if your view is different

(Dinesh Sairam) #642

That’s exactly why I’m asking you to share your notes. Your method is obviously different from mine. That’s why I’d like to see where you stand when it comes to valuing DMart…

If you’re busy, then fair enough. I only asked because you said you have an automated tool in Screener, which shouldn’t take much time at all.

(Bheeshma Sanghani) #643

Hi @edwardlobo

Not sitting in judgement here!

All i am saying is that if you give a company to value to a bunch of people that you trust and if they all come up with widely differing estimates then one should conclude that valuation is pointless as there is too much estimate error. If on the other hand if there is reasonably small gap in estimates , then there is a purpose to valuation.

My comment is not on the methodology but on the variance of the estimate provided by the methodology. For e.g Dinesh has come up with an estimate of Rs 1168 for dmart.

I know of several individuals who have also valued Dmart independently ( all smart and proficient in valuation) and have come up with estimates ranging from Rs 500 to Rs 3000 using DCF. This kind of defeats the purpose of valuing cos using DCF in my view. Maybe there is something wrong with the tool, the assumptions or the analyst. I dont know. But variance cant be this much. That’s all i am saying.

(Dinesh Sairam) #644

I’d like to interject by saying that variance can be that much. Opinions form markets, right?

Value is created from cash flows, the growth in cash flows and the required discounting of those cash flows. Any difference in any of these 3 areas between Analysts will create a variance.

Ultimately, with a stock, you’re buying a story. You’re trying to attach numbers to that story. If the numbers give you a value that’s within your comfort zone, you think about buying it. If not, you don’t. Why beat yourself up over what 1000 people think about the company?

For academic purposes, that’s actually fine. But when my own money is in question, I only listen to myself / question my own assumptions.

(edwardlobo) #645

I agree @bheeshma, sorry to have misunderstood

Basically if you take all capital of Dmart and put it in an Fd at the end of 20 years you’d get say 500k whereas their business turns it into 700k plus the interim cash flows that I haven’t counted

Dcf Valuation is not an accurate tool so if the present value of invested capital is more than sp I don’t see a point in trying to arrive at an accurate value for an impossible figure

(edwardlobo) #646

I can upload the spreadsheet but it’s back of the envelope just numbers
I don’t even mention the name of the company
It pulls last profit from screener, and has formula for npv and terminal value
The growth percentage is a backsolver from excel
What it does it for the current valuation based on cost of capital being roughly 18pc what growth is required and over what period
I then decide if it is possible for the company to surpass this growth and by roughly how much. This part is down to individual assumption
My required growth for current valuation came to 13pc
If it grows at 30pc, I see a large gap
I’ll send you the spreadsheet in your email but it doesn’t have any narrative
I have made scripts that went through each company on screeener with a time delay so I don’t overload the server but I could use numbers from any website that hosts them
This gives growth required for all companies
Some have zero or less like thirumalai

Markets can take for ever to notice the gap so I then research into these companies to see if there is some event like expansion that will bring this difference into spotlight