Dinesh Sairam's Portfolio: Requesting Feedback

(Bheeshma Sanghani, PhD) #183

In all cos that engage in a low margin high leverage business and financial cos are a special case of that, there is typically a significant management premium inbuilt into the price over and above the other risk premiums. Otherwise govt owned financial cos are every bit as good (or bad) as privately owned ones from a business point of view.

In uncertain times, this premium gets eroded and intangible assets once eroded have a economic replacement cost which is far in excess of its original cost unlike tangible assets which can be replaced in real terms.

I think that when investing in low margin enterprises one should pay heed to unusual movements in the stock price. This is not applicable to other more easy to understand and more tangible businesses where an intrinsic value can be assigned with ease and there is no problem with buying more if the price falls.

I would suggest one draws a long term trendline on the price chart touching at least 3 lows when investing in financial cos and any break below it should tell you what you need to do or else one should invest in govt owned enterprises which have no “management premium” to begin with so the question of erosion doesn’t arise in the first place.

Just my views

(Changu Mangu) #184

Dear @dineshssairam

I read your blog and I just have one bit to share with you.

I think your skills to value the business on pure financial metrics are quite good. But I feel you are not paying any value to Business Quality & Management Quality.

If the entire theses is a financial valuation then I think you will end up with a large collection of businesses that are value buys only on paper (DHFL, Cupid etc). Then when some further adverse event happens you have a business that only becomes cheaper. Would you consider thinking why a ITC or other such stocks get a premium? Could it be MQ and BQ should also have a weight. I think it is fine to have 10 or 20 percent in value plays but with your methodology would you not end up with a portfolio of probably (not sure) junk stocks only. Yes, they may pay off, but is that the best use of one’s skills.

I think if you also made the effort to give a weightage to BQ and MQ then your valuation will make more sense. We should not be judging a business only by it’s numbers. The soft factors should have a strong weight as well.

Just my 2 cents but I wanted to make sure I send this message to you.


(Dinesh Sairam) #185

Thank you for the message.

But in my humble opinion, ‘Business Quality’ and ‘Management Quality’ aren’t magical things that automatically warrant overpaying. One should think about how these will affect the value of a company.

Take a company like Maruti. Their business is amazing because they’ve managed to build a strong distribution network. Their products can reach far and wide as soon as they’re introduced. But should this mean that you automatically pay a 20% premium while purchasing Maruti stock? Obviously not, at least that’s what I think. You should think about how this vast network will help their business and how their numbers are likely to be based on that advantage. This can then translate into additional value for the company’s stock.

The bottomline is, whenever you’re thinking about the price/value of a company, you’re automatically making assumptions and doing projections. It’s just a question of whether you’re doing it deliberately or unwittingly.


Hi Dinesh, I too have gone through your blog and your analytical skills are top notch. Commenting on those lines is beyond my scope… Period :pray:

I am not in favour of hero worship either, but MQ & BQ have a big impact. Overpaying is justified. I would also add KQ. Knowledge quotient. Companies with great MQ, BQ and KQ has better moat, better margin of safety, attracts FII and lets not forget ethics as a cornerstone to long term sustainibilty of business. These are extremely qualitative and if at all it needs to be quantified, then one may add 10% premium on each factor. I read somewhere in VP only, Indian investors typically just don’t value R&D capabilities. Its like a cost line item nothing more…Can’t imagine such broker reports valuing Google, Tesla, Microsoft

(Roy) #187

DCF Valuation is all about estimating future cashflows and then calculating the present value of these estimated future cash flows.

I believe Dinesh’s ‘numbers and narratives’ model already takes into account the quality of management, quality of the business, R&D spends etc. Ultimately, each of these factors affect the estimated future cashflows and hence the fair price of the stock.

Assigning a further 10% premium because of better quality of management, etc does not make sense because these have already been factored into in arriving at the fair price of the stock.

(Divyanshu Bagga) #188

The validity of any model is given by the quality of its predictions. Let’s roll back three years, and apply the model to Gruh and DHFL. Which one is cheaper? If the answer is DHFL, then I will call the model flawed. I do believe that the stock price reflects the discounted sum of all future cash earnings, so there must have been some mistake in the growth profile we have assumed for these business. It is to compensate for this mistake that we add quality premium or subtract corporate mis-governance penalty. But ultimately it is not a premium. We are not overpaying, rather we were able to buy Gruh “cheap” in order to generate above average returns.

(Dinesh Sairam) #189

Yes… in a nutshell, this is what I intended to say. If you think “Quality” counts, think about how it will affect the business. I just don’t think considering “Quality” a magical phenomenon floating in the air is justified.

(Dinesh Sairam) #190

My model accounts for R&D, Market Yields (Debt), Operating Lease payments and Employee Stock Options.

The only thing that I ‘approximate’ in my model is the Value of Subsidiaries (I suggest the usage of a Multiples Valuation instead). But then again, you can Value the subsidiaries separately in a DCF and plug in the value.

([email protected]) #191

I had a very insightful offline discussion with dinesh when I was trying to back test his model on Page. I was trying to figure out what Page would be worth based on his model in 2013. Based on a high level assement by Dinesh and a bit detailed by me, we independently arrived at a price which was 5% undervalued (by me) and 7% overvauled (by Dinesh) vs current market price. Given this small margin, it all came down to managment quality, competetive opportunities in market and personal risk profile & conviction in the model. Note that the price of the company had gone up more than 70% every year between 2010 and 2013. Also it was trading at 35 PE, at a time when there we a lot of value buys around (HUL at 27PE or TVS Motor at 14PE that I bought)

To test what the model says now, I used it for current year estimate and arrived at value at which the current price is 4.5X. Attached both Models.

For me the premium between the two valuation reflects the confidence (or irrational exubarnace, time will tell) of what the management can achieve in the next 5-10 years, given what they acheived in the past.

This is a singular test thus I cannot draw a lot of conclusion from it. However given my experience in the market, I do not ignore the potential of a good managment (Model´s are supposed to guide us and not lead us), but at the same time generic % don´t help either. So I treat each investment at their own merit and decide the margin of confindence respectively ( for me its the opposite of margin of safety).

If anyone is interested, I can share a copy of above mentioned conversation with Dinesh.Numbers and Narratives (Compact) - 2018.xlsx (24.1 KB)
Numbers and Narratives (Compact) - 2013.xlsx (25.0 KB)

(Shailesh) #192

I just happen to go through your discussion with peers on DHFL . My two cents

Each one has investing process and I see you have very strong conviction in the same , Don’t let others shake it …

In long run process however rudimentary it is - If you stick to it in bull and bear market you will beat 95% of investors …

Best wishes …

(EL) #193

In your model a stock is a function of all future dividends multiplied or by some form of (discount rate and growth rate) percent to arrive at ongoing dividends and terminal value.

Is it possible to start with present stock value and reverse calculate to see how much growth rate is assumed into the present stock value ? So you don’t have to guess what is below or over your price but understand what growth the market is pricing and if the company were to throw enough resources to attract suburban markets if that growth can be achieved.

Secondly the discount rate is a function of “expected” risk free rate plus the risk percentage. In theory I think risk cant be same for Page and Colgate or Page and Kitex Garments

If a company can grow on debt and debt is 7pc whereas the discount rate is 12pc over the long run any company run on debt, albeit risky, will look more promising. Although we have increased risk by increasing loan but most models don’t address this. Ofcourse in the end valuation you do remove the debt amount but its not the same as removing the growth and risk that has come from that debt.

Extending the discount rate subject a bit more, company stocks together with real estate, gold, bond and bank fd compete with each other for investments. If there is a deep value discount in one market, money usually flows into those markets.

If we are evaluating 2 fixed interest fds, do we evaluate interest of one bank with 2013 rates and interest rate of another at 2018 rates or do we adjust those to current conditions ?

So 2013 and 2018 discount rates probably are considerably different

Disc: Not invested in Page

(Dinesh Sairam) #194

You can probably tinker with the growth inputs until the value equals the current price. But that defeats the purpose of the model. You don’t look at the numbers and then try to build a story from there. It’s backwards. We value a company as we approach it - first the story, then the numbers, then a price to pay for both of them.

But in short, yes, it can be done.

They’re not. Theory defines Standard Deviation as Risk. Standard Deviations are usually higher for companies with unstable earnings and lower for the stable ones.

I personally don’t agree that Standard Deviations is Risk, although I’m forced to use it for my public valuations (For lack of a better alternative benchmark). Otherwise, I use my Cost of Capital to discount and then I do a Monte Carlo Simulation of the different future projections, so as to determine the probability levels at each given price point (i.e. Probability of Undervaluation/Overvaluation).

This, co-incidentally is what Warren Buffet also does. He does not adjust for Risk in the Discounting Rate. He adjusts for Risk in the Cash Flows (Although, he is very well capable of doing a quick calculation of the probabilities in his head). More of my thoughts on the Cost of Capital here:

So, naturally, companies with stable growth and returns like Page won’t differ wildly from their initial calculated value (Say, if my initial calculation shows a value of Rs. 20,000, the extremely pessimistic and extremely optimistic ranges may well be between Rs. 18,000 and Rs. 22,000). Whereas for a company with highly unstable earnings and returns (Say, Kitex), the spread might be big. You get to decide at which probability level you will pull the trigger. Perhaps for Page, you would require only a 60% Probability of Undervaluation, but for Kitex, you might need 95% – this is left entirely to you.

Models do do this, or should I say, the market does? Beta, an inherent component of the Cost of Capital, is higher for companies with higher Debt. Subsequently, Cost of Debt, also a part of Cost of Capital, is also higher for companies with higher Debt.

In fact, I don’t directly use the Market Implied Betas for calculating the Cost of Capital. I use the Bottom-up Beta, which comes from comparing the specific company’s financial risk with that of the industry. Put shortly, if the company in question has a higher debt than the industry, it will end up having a higher Cost of Capital (Lower PV of Cash Flows) than the industry. Operating Leverage can also be added to this process, but it’s difficult to get the split between Fixed and Variables expenses for many companies in an industry. But you get the idea.

Yes. The Risk-free Interest Rate forms the basis of the Cost of Capital of every single asset in an economy. As you mentioned, when Bank Interest Rates change, they don’t do so in a vacuum. They change because the RBI has guided them to – via the Repo Rate, which over the long term becomes the Risk-free Rate.

As I’ve mentioned elsewhere: In the economy of 0% interest rates, I will be happy earning 5% RoI. In the economy of 8% interest rates, I will be happy earning a 15% RoI. It’s all about Opportunity Cost or as Aesop said “The bird in the hand is worth two in the bush.”

I hope this made sense.

(EL) #195


It does make literal sense mate

Beta is a variance between stock price performance compared to market as a whole ?

(Dinesh Sairam) #196

Yes. Beta is how the stock reacts to a move in the market. A Beta of 0.5 indicates that if the market rises by 1%, the stock will rise by 0.5% and if the market falls by 1%, the stock will fall by 0.5% and vice versa for a Beta > 1. Stocks with a Beta > 1 are termed ‘Risky’ in Finance Theory.



You are a finance geek :pray:

Please help me with one metric. How can I find companies with high operating leverage. Give some simple examples.

Why I asked is, if a company has high op lvg, then profit can rise much faster than Sales…that is if sales grows by 10%, profit can grow by 18%…

How will I use it.
I will look for companies, small mid caps, which can grow sales by 10 to 15 % for next 3 years, however with better op lv, the net profit will rise faster… So will stock price.


(EL) #198

So using A market behaviour in the formula we are crediting market for having correctely priced the stock so far and also at the same time hoping our formula will show companies that are trading at a discount.
Another thing is by giving importance to the ratio in our formula are we are expecting the stock will continue to perform in the same way it has in the past with respect to the whole market ?

(Dinesh Sairam) #199

Saying that estimating the Discounting Rate is ‘pricing the stock’ is a big stretch. It’s just one part of the exercise. In fact, it’s the expendable part. Everyone has different ideas of Risk, and hence, different Discounting Rates. You may require only 10% from something, whereas I may require 15% and so, you will be willing to buy more things than me. And that’s completely fine.

Once again, I suggest you read my article of Cost of Capital.

(Dinesh Sairam) #200

Absolutely, you got it right. Way earlier in this thread, I have mentioned that I use a Ratio called the ‘Profit Growth to Sales Growth’ Ratio (A custom Ratio I came up with for my own purposes). I built the Ratio is Screener using the following formula:

((Profit growth 10Years / Sales growth 10Years) + (Profit growth 7Years/Sales growth 7Years) + (Profit growth 5Years / Sales growth 5Years) + (Profit growth 3Years/Sales growth 3Years)) / 4

I personally wouldn’t pay too much attention to it, though. Operating Leverage is a double edged sword. In an inflationary trend in the industry (Increasing Cost of Operations), you expend in historical cost (Lower historical cost) and earn in current revenues (Higher current revenues). In a deflationary trend in the industry (Decreasing Cost of Operations), you expend in historical cost (Higher historical cost) and earn in current revenues (Lower current revenues).

The solution to this, would be to find companies with decent pricing power. They can make full use of the Operating Leverage in an inflationary industry and manage to cap the damage during a deflationary industry trend.

This is why plain vanilla commodity types businesses with huge Operating Leverage (Like Oil giants, for instance) struggle to create value for the investors. The economics of their industry is in a flux, but they aren’t as nimble to adapt, because their Fixed Expenses are making them slower.

(Mahendra243) #201

Hi Dinesh…Need your view on the ambit HFC report…i didnt get hold of the ambit report so posted a link to twitter handle…pls advise

(Dinesh Sairam) #202

MCLR or the Marginal Cost of Lending is calculated with the Repo Rate as the base. In short, Repo Rate + the Operational Costs of the entity + some Margin will become the MCLR for the particular entity. I mean, the earlier (Base Rate) calculations were also along similar lines, but they were less strict and less regulated by the RBI.

Banks have already come under the purview of the MCLR since 2 or 3 years back, but they have done quite well until now, haven’t they? There were also similar “doomsday” type of articles when MCLR was introduced by Mr. Rajan. Here’s a sample from 2016:

In essence, I think this is a much welcome change. NBFCs will now think about how they lend, rather than mindlessly chasing growth. In fact, this whole phenomenon is being painted red because growth is going to drop. But most NBFCs have already dropped due to the crisis. The market has already priced in the slowdown in growth and then some more. The growth going forward will be more responsible and that’s fine by me.

From a more theoretical perspective, growth rates will drop, but Cost of Capital will decrease because the NBFCs are now being closely monitored by the RBI. In the end, as with any investment, it boils down to the Risk-Reward trade off. Would you rather have a company growing at 30% that’s very risky or a company growing at 16% which is very safe? Or somewhere between the two?