Thursday, September 11, 2008

Perils of PE

For avid readers of my blog, you might have realised that I am undergoing a deconstruction process of my valuation techniques. Just yesterday, I was debunking my own way of using a bastardized version of DCF to calculate the intrinsic value. Mr.Investor could very well describe me.

It's important to read about investing and it's also very important to practice it by valuating listed companies. Only then will the things read be transferred from the RAM of your brains to the more permanent hardisk. For now, I'll be deconstructing my PE ratio analysis.

We can use the PE ratio to derive the future price of a company, if we also know about the earnings in the future. To go about doing that, we just need to follow the steps below:

1. Assume an earnings growth rate. Project future earnings by applying the earnings growth rate to current earnings.

2. Assume a PE ratio. Then multiply the projected future earnings with this PE ratio, and we'll get the future price.

For example, company A has earnings growth rate of 21.1% (CAGR over 4 years). For simplicity, we assume a earnings growth of 20%. FY07's EPS is 5.6 cts per share. Thus, projected FY08's EPS is 6.7 cts per share (5.6 x 1.2 = 6.7).

Here's the historical PE ratio:

* EPS in FY07 is actually 0.278 RMB, not 0.278 SGD. There is a mistake in the headings.

The lowest PE ratio is 5.5x and the highest is 26.3x. To be conservative, we assume a PE ratio of 5.5x to multiply the FY08's earnings of 6.7cts per share, giving us a future price of $0.365.

Easy right? But there is actually a lot of things hidden beneath the simplicity of the maths. Besides the difficulty of assuming the earning growth rate, there is also the problem of using historical low PE ratio to justify the multiples. Actually these two are inherently the same problem - that of projecting the past into the future.

To illustrate, imagine this. A 90 years old man, having lived to such a ripe old age, had never died before (obviously!) in this entire lifetime. Based on historical data, this old man will continue to grow at the CAGR (based on the past 90 yrs) of 1 year per annum. In fact, he will continue growing for the next 90 years, based again on past data. If we can extend this further, he will probably never die because past data had no records of him dying before. What do you think of this?

To assume that the price of the company will not fall below the historical low is a myth, and is best dismissed with facts. In 2003, if we are to compile the historical PE data of company A, the lowest PE is 8.5x. If the investor bought in at 8.5x, thinking that the price will be supported at that PE level, he will be sorely disappointed. In 2004, lowest PE becomes 7.1 x. In 2005 and 2006, lowest PE becomes 5.5x. The lowest historical PE just keeps getting lower and lower.

Based on 1H08, EPS for Company A is 3.126 cts per share. Current price now is 24.5 cts. If we annualised 1H08 by multiplying by 2 (assuming constant earnings for 2H08, haha, the irony of it), we end up with a FY08 EPS of 6.252 cts per share. So guess what's the PE? It's 3.9x.

PE can get lower than what past historical data showed. To project the past into the future, is akin to a 90 year old man projecting his future life span forever because he had never died before. Amusing but ultimately ridiculous.


DanielXX said...

hi nice article. Incidentally I had also written something along this line in a recent blog article. You could check it out at Mystockthoughts blog.

PE has caused grief to many in this bull-to-bear market. Now it's morphed to premium-to-book or even book valuation. Tens of billions of justifiable market cap have simply evaporated through such a change in valuation.

la papillion said...

Hi danielxx,

Yes, i've read your article :) It's very insightful, detailing the different valuation technique to showcase the best. Different valuation metric sells at different times, in era, those companies without earnings will have Price/sales ratio or eyeballing hits. Way before all this, where stocks are deemed risky and bonds are the norm, the metric to sell stocks are dividend yields.

So now, when all hope is lost, stocks are sold at having low price to book :)

There are seldom new things in the market, isn't it? :)

Anonymous said...

nice article there;) p/e valuation is one of the more dangerous i believe. Re-ratings occur all the time, and I like the last statement u made. The best I believe might be to use the book value, which probably provides u with some effective protection from downside during trying, and also bullish periods.


la papillion said...

Hi patrick,

Thanks! I have some comments waiting to be typed out in your blog too...need to find the time. U're interested in 2nd chance? :)

Mike Dirnt said...

Hi LP,

PE when used alone may not be good enough.

how about calculating the PEG ratio? that will give a better sense

by the way, i just bought into Pac Andes today. NTA is at $0.23 and curent price is slightly below it. so quite reasonable enough for me. i did not use NAV because it has a big intangible assets in it.

i dont think i can catch the bottom, but im looking long term for this one

la papillion said...

Hi mike,

Ya, PEG by Peter lynch can be a good guide. I think the safe limit to start looking is when the PEG hits around 0.5 - that means a 1% earnings growth rate is valuated at 0.5 PE.

But PEG doesn't work on cyclicals. For cyclicals, the best time to buy might be at high PE, signalling the bottom of the cycle (peter lynch mentioned that, not me).

Pac andes have very good dividends yields at your price, if it's sustainable :)

Mike Dirnt said...

Hi LP,

I did not come across about Peter Lynch high PE. That is interesting to signal a bottom. Im scratching my head now