Tuesday, January 26, 2010

A Two Iteration Monte Carlo Simulation on Trading

This is something that I have posted in a comment some time back. I thought I would just expand it for discussion and see if it makes sense.

First let’s work through some assumptions and no.s and see what’s the expected return for trading.

1. The capital base is $100,000
2. $10,000 is utilized per trade
3. 10 trades is done in 1 year
4. Take profit at 20%
5. Cut loss at -10%
6. Winning rate 60%
7. Transaction cost $20

Based on these:

a. The 6 winning trades will bring in $12,000.
b. The 4 losing trades take away $4,000.
c. Transaction cost is $400.
d. Total winnings: $3800
e. Return 3.8% - Yeah that's life for a trader, my darling. Why don't you put the money in CPF and earn the same return?

Ok, there are a lot of assumptions, some might be skewed to put traders down. After all, this is a value investing blog. :) What if we tweak them around? Say the capital base is just $20,000 – then the return becomes 20%! However the rationale would then be it won’t be possible to realize 10 trades in 1 yrs with just $20,000.

Anyways let’s do a more aggressive one

1. The capital base is $50,000
2. $10,000 is utilized per trade
3. 10 trades is done in 1 year
4. Take profit at 15% (rationale being that the time horizon is now shortened)
5. Cut loss at -10%
6. Winning rate 60%
7. Transaction cost $20

Based on these

a. 6 winning trades will bring in $9,000
b. 4 losing trades take away $4,000
c. Transaction cost $400
d. Total winnings: $4,600
e. Return 9.2%

Ok that’s better than market return, but that’s probably also a high hurdle. To do 10 trades with $50k in 1 yr, reach trade can only go for 6 mths.

I think the appropriate scientific experiment we should do is a Monte Carlo simulation of 1,000 iterations to see what’s the true expected return. But my guess is it’s actually going to be less than market return (of 8% or so). Yes, actually if you do it correctly, a trader should earn positive return, not negative ones. And in all those books, it always says academic studies show that trading cannot beat market return after factoring in transaction costs.

Well this also implicitly means, if you get your transaction costs low enough, you might beat market return and becoming a Big Swinging Dick.

Ok, daydreaming over. Trading is hard. My sense is, it is actually much harder than value investing. If you do it right, you might just make average return. Most people don't do it right in the first few years. Think about the time and effort that is needed to execute these trades during the year. Basically it’s a full time job in itself. Not forgetting that it's gonna be one helluva emotional rollercoaster ride every day!

Well, that’s why I stick with value investing.

Thursday, January 21, 2010

Management compensation

Needless to say, if the management team is paying themselves too well, pls avoid the company. Most annual reports of Singapore co.s these days have a section on management payout. I make it a point to find out how much they are paid.

Just some rough no.s (since I can’t really remember all the figures), the CEO pay package is usually about $1mn for a few hundred million revenue firm. For smaller co.s, it is about $500k or so. Of course, as we all know, the record is a whopping S$20mn.

What is a good sum to pay a CEO? And how to actually determine the formula for the payout? Well I don’t have a good answer, but what I do think is wrong is to base it off revenue. Bcos a firm could have high revenue but zero profits to shareholders. It is also wrong for it to be mechanical, like based on formulas. So maybe a basic package and then bonus to be based off a comination of factors like net profit growth, impact of past decisions and qualitative appraisal by stakeholders of the firm.

Well there is also the social pay scale to consider, in our crazy world where a 23 yr old analyst could be paid S$100k a year, surely we cannot expect CEOs to be paid like S$150k a year right? So actually there is a floor for CEO’s pay. Since senior managers in big firms get around $200-300k so it is not unusual for CEOs to be getting around $500k at least.

Most of the time, when reading the annual reports, you won’t find anything strange until it gets out in the news. Usually the annual report just says that top management is being paid in a range of S$1-5mn, which is reasonable, considering what we have discussed.

Strange things happen once in a while and astute investors’ warning bells should sound and put companies that pay their CEOs or top management too well on the blacklist.

The infamous case of Sing Power comes into mind. I cannot recall the whole story but apparently the compensation package for the top management exceeded the net profits of the firm or something. This was bcos is was based on some arcane formulation and the management argued that it was ok. My foot!

Noble group made the news paying 11 directors $30+mn in 2008. Not sure if this is a lot or not. Net profit was a record $500mn or so. So maybe it’s reasonable. After all, only 5% of net profits right? But I checked out their dividend payment – it was also $30+mn. Hmmm...

Of course we always have our favourite CEO who was paid $20mn – highest paid CEO ever in the history of Singapore in a year when his firm profits was down 50%. Again the formula excuse was used to justify this absurdity.

The lowest paid CEO in a Fortune 500 firm, by the way, is our favourite hero from Omaha. He pays himself US$100k annually.

Tuesday, January 12, 2010

More on Dilution

Rights issue sucks! Let’s see how this works:

Imagine you are the sole owner of Company ABC. It's better to think as a sole owner, it makes things clear.

So you put in $10mn capital to start the co. You list the co. and now own 1mn shares of $10 each. ie mkt cap of your co. is $10mn. Then you appointed a CEO to help you run the business. He lost $8mn, well partly bcos of the crisis, partly bcos he was not prudent and expanded to rapidly during the heydays of 2006-07, partly bcos he paid himself and his kakis $1mn over the past few yrs.

The share price plummet to $3, ie mkt cap is now $3mn and the capital is now $2mn. So the CEO announces a rights issue of $10mn, 5mn shares to be issue at $2, that’s 33% discount to today’s price of $3. Do you:

A. Rejoice bcos now you will own 6mn shares of your co. at an average of price $2+ but the share price is $3 (btw you paid $20mn in total, but the value of your co. is now only $12mn)

B. Or you curse the manager for losing most your capital, fire him and sue him in court to try to salvage part of the lost $8mn.

Rights issue is a form of dilution: if you do not take up the issue, your stake in the co. is reduced. If you do, you just gave money to a crappy management who lost the original capital in the first place.

Some other companies do outright secondary offering where the original investors suffer if they do not participate.

Management will often say that raising capital is the prudent thing to do to keep the company as a going concern. But who jeopardized the firm as a going concern in the first place?

Of course, in the stock market, where there are a million participants, Genius Ah Beng could have waited for the shares to fall to $3, participate in the rights issue, bring down his buying price to $2+ hence making a arbitrage since today’s share price is $3.

But that does not change the fact that management screwed up in the first place.

If you are a value investor, you should not be giving money to a management who goes cap in hand whenever he gets a chance! (And it’s always a HE, not a she). The SHE gives out the money magnanimously, every time!

So pls beware of companies that to serial rights issues!