Saturday, July 07, 2012

101 on Investing

Three points to note about investment advice that generally floats around:
  1. Market movements almost never mirror reality/logic, contrary to what theories such as “Efficient Market Hypothesis” will tell you. As Keynes once remarked “Markets can remain irrational far longer than you or I can remain solvent
  2. Analysis always tends to be post facto. Therefore, with the wisdom of hindsight, past events may be justified. But there’s no guarantee that future events will play out similarly.
  3. Most brokers (in the garb of ‘experts’) are paid to peddle stocks on business news channels by certain investors. Use your judgement before committing capital based on hearsay!
We invest in equities to make superior returns on our capital (vs fixed deposits and bonds). There’s greater risk obviously, but we hope our insights/information/luck will help us get to those superior returns. Aim of this post is to analytically think through what the drivers of those returns are and how the math behind it works.

Drivers of Returns:

There are four aspects that primarily move the needle on returns, and all of our investment diligence should therefore be centered on these issues:
1) Growth: Splits out into market growth and market share. Say a market is growing with India’s nominal GDP (15%), and you think the company can gain some share over time (a nuanced sub-segment level understanding of each of the product lines of the company is necessary to make this claim!) You assume that the share gain will be rapid and you underwrite that it would outpace the market by say four points. Implicitly your revenue growth assumption is 20% (16%+4%).
In the ideal world, you want to find a company that has some of the following characteristics (obviously non-exhaustive) to give you comfort about growth & share:
  • Strong competitive position in a high-growth industry driven by high barriers to entry (say technology / patents) and brand loyalty (e.g., iPod)
  • Scalable business in a concentrated industry that would see a small subset of players capturing an unfair share of the market growth going forward (e.g., 2 wheelers)
  • Relatively free from regulatory hurdles and govt/political influence (therefore keep off infrastructure, real-estate etc)
  • Low customer churn with long-term sticky contracts (e.g., BPO) or low customer concentration (e.g., retail industry)
2) Margin: You have to take a call on COGS (Cost of Goods Sold i.e. direct costs attributable to the production of goods or raw material costs +conversion/direct costs) and SG&A (Selling, General & Admin expenses). These two are your main cost buckets that flow through to your EBITDA.
EBITDA stands for Earnings before Interest, Tax, Depreciation & Amortization. When people talk about “operating margin”, they refer to EBITDA margin. EBITDA = Revenue – COGS – SG&A. Other costs below EBITDA are depreciation & amortization (both are non-cash expenses), interest on your debt (if any) and taxes.  EBITDA – Depreciation & Amortization – Interest – Tax = Profit after Tax or PAT.
Now that we have the definitions out of the way, you should pay close attention to how you think gross margins (Revenue less COGS) will play out in your investment period. Say you assume it will improve due to price increase (basic inflation + price premium for your products due to better perceived quality) and lower conversion costs. Say you assume you would gain 0.5% of revenue each year on gross margins on account of these two levers. 
You also have to consider SG&A – you could get some leverage from employee expenses, better utilization of facilities etc but could consider increasing spend on sales & marketing. In our case, let us assume all the savings from the former is invested in the latter (i.e. sales) and your SG&A margin remains constant every year.
Therefore, in this case, your EBITDA margin will expand by 0.5% every year (0.5% increase due to gross margin expansion and no change in SG&A) 

3) Multiple: The multiple on EBITDA or PAT is technically meant to be the closest proxy for future cash flows of the company. In reality though, it changes every minute as expectations change depending on macro environment, regulatory changes, liquidity, significant events at other companies etc. To evaluate the “right” multiple to pay, it’s probably wise to compare the current multiple to the five year average to get a sense on how the multiple has moved over time. In the ideal case, you don’t want to be buying off a high. Remember the age-old truism “Buy low, sell high”.
Super-normal returns can be made if you believe that the market is currently under-valuing the company and you have sound reason to believe you are seeing something that the market isn’t and that the market will hopefully come around over time. Many industries have re-rated upwards (e.g., consumer) and many have re-rated downwards (e.g., Mid-tier IT) over time, so it is crucial that you see today’s multiple in perspective and take a call suitably.
In general, market-leaders trade at a slight premium, so if you believe your company (say currently at #2) will eventually be #1 due to superior product innovation, better management etc, you could underwrite a click or two of multiple expansion. Nonetheless, this is one area where you would rather be conservative and wait to buy when there’s a steep correction due to an uncorrelated event.

4) Changes in capital structure: This is a slightly more complex topic and deals with capital structure of a company. It is rarely a driver of returns in India (given high interest costs, regulations etc) so I’m not in favor of a lengthy discussion. But here’s the short of it, for those who are interested:
Say a company in the US is worth $1B and it’s debt to equity ratio is 4:1 i.e. $800M of debt and $200M of equity. You use the company’s current cash flows to pay down say 50% of the debt in your investment horizon. Even if the company’s enterprise value remains constant and nothing else in the business changes materially, you are now left with $400M of debt => $600M of equity i.e. 3x return on your equity simply via debt pay down :) This math gave birth to the LBO industry.
Let us put what we have learned to good use and do some simple math.

Basic math on how returns work:

Say a company has year-end revenues of Rs. 100, EBITDA % of 20% and PAT % of 10%. 
For the sake of simplicity, let us value the company on LTM (last twelve months) Price to Earnings (PE) multiple and say the multiple is 15x. Market cap of the company = 15x PAT = 15 * 10% * 100 = Rs. 150
Now say there are 15 shares of the company. Therefore, each is valued at Rs. 10. You buy say 5 shares from an existing shareholder (technically a ‘secondary purchase’). Now you own 5/15 shares i.e. 33.3% of the company.
Say five years down the line (typical time-horizon for a long–term investor), the company has grown its revenues at 20% CAGR and has been able to maintain EBITDA and PAT margins (over-simplistic, but let us go with this). Assuming the multiple doesn’t change, you would make 20% returns YoY on your 5 shares. Your Rs. 50 now is worth 50*(1+20%)^5 = Rs. 124. Therefore, you have a Multiple of Money (MoM) of 2.5x and an IRR of 20%. Simple?
Let us flex the other variables as well, so everyone has the math clear in their heads:
Say the company’s revenues grew at 20% YoY. Let us now say that the company did a bunch of operational improvements thereby enhancing its margins and say its PAT grew by 25% YoY. Also assume that the market starts loving the company and starts paying a premium on its shares, now the multiple expanded from 15x to 18x. What returns will you make on your 5 shares purchased @ Rs. 10 each?
The simple P&L will look something like this:

Therefore the company’s PAT at exit is Rs. 31, the multiple at exit is 18x, therefore market cap at exit = Rs. 31 * 18 = Rs. 558, up from Rs.  150 at entry.  Share price is now Rs. 558/15 = Rs. 37.2. You own 5 shares, hence your shares are now worth Rs. 37.2*5 = Rs. 186
Therefore, your simple returns are:
MoM = 186/50 = 3.7x
IRR = 3.7^(1/5)-1 = 30%. Neat! :)
Note: This return math assumes none of the following have happened in the investment horizon - dividends/re-caps from excess cash accrued, management options, interim primary infusion, back-leverage and all other second order issues


I will (hopefully soon) come up with a separate post on “Nuances to investing in India” where I will talk about promoter/governance issues, issues with ‘control transactions’ in India, how bankers try and raise expectations etc.
If you’ve reached this part, I am assuming you read the post and I hope this was helpful. I have tried to keep it simple, but if you have any questions, I’m happy to answer them. You can reach me at utsavmitra@hotmail.com