Steel Industry

After 186 posts about value investment philosophy, I think it’s about time to write about something else. Well, after all, value philosophy can actually be surmised into just 3 words. So, I am actually quite amazed why I could write so much. So going forward, hopefully I can write about industries and individual stocks. As and when new ideas hit, I will still talk about value philosophy and the big picture. Ultimately, that is what’s most important and what will drive long term return for investors.

In this post, I would like to talk about the steel industry. Steel is a basic commodity used by humans and has been pretty integral throughout the development of our civilization. Sadly as a business, it sucks. The industry as a whole doesn’t really create much value for shareholders although there are periods where it churns out enough cash to whet some appetite.

Today, about 1 billion ton of steel is consumed every year. China accounts for half of the usage. Outside of China, Asia including Japan, accounts for bulk of the rest. Well, this is unsurprising as steel is mostly used in construction and infrastructure which Asia needs, a lot.

The business model is simple enough. Buy raw materials like iron ore and coking coal, throw it into a blast furnace, out comes molten steel, add some other metal to make it better (like nickel for stainless, or zinc coat it for shine) and process it into sheets or beams etc. This in itself is not bad. What is bad is:

1. Both the input and output prices are uncontrollable.
2. Competition is very, very tough
3. It is very capital intensive

Raw material prices are controlled by the ore majors: BHP, Rio Tinto and Vale. Specifically, they dominate the spot market and use the spot prices to determine contract pricing. So the steel makers have no say in pricing. The final product prices are also determined by the spot market. There are international market prices for a variety of steel products including the most famous hot rolled coil (or HRC), for H-beams used in construction, for pipes etc.

The reason why such spot markets developed is probably bcos there are simply so many players in the market that is just have to be done for the benefit of both the steelmakers and their buyers. With such markets, products could be standardized, distributors can handle them easily and lengthy negotiations could be avoided. But that’s bad for profits.

But why are there so many steelmakers globally? Well, in the past, it was a country’s ambition to have its own steel mill. It’s a symbol of strength for the nation. The western countries had it. Japan still has it. Korean has it and now China and India are building theirs. What’s worse is when the various provinces or prefectures also decided that they should have, hence you have all these few hundred steelmakers all over the world, each having less than 1% of the global market.

In the middle of this decade, someone decided to restructure the whole industry. His name was Lakshmi Mittal. So he started buying small steel mills all over the world. But he realized that wasn’t enough. There were just too many. In a move that shocked the industry, he decided to buy over one of the biggest steel players globally. Today, his company is called ArcelorMittal and it has capacity of 100mn tonnes or 10% of the market.

But still, 10% is nothing in a world where the suppliers and customers are much stronger and you still have over a few hundred competitors. ArcelorMittal, amazingly, has been able to generate good cashflow by squeezing cost and investment. Unfortunately, the money has to be used to pay down debt and it will take another 5-6 years to bring debt down to a comfortable level. Not to forget, by that time, it probably needs to resume its capex plans as well.

Which brings us to the 3rd point. Steel is insanely capital intensive. It takes USD 1,000 to bring 1 ton of new capacity on. For ArcelorMittal to increase capacity by 10%, it will cost USD 10bn! That’s one sixth of its equity base today. Most other steelmakers are not even that half its size and a new blast furnace project almost always means new financing.

So in short, the steel business, though integral to the development of our civilization, is bad business. There is usually nothing left for shareholders, after everything is said and done.

Well, that is the big picture. Value investors are also stock pickers and hence the dynamics can change for individual companies.

Buffett had a stake in the Korean steelmaker POSCO for the longest time. The story for POSCO is that the company is the No.1 leader in a country that is perpetually in short of steel despite being one of the biggest exporters of steel intensive products like ships, cars, and consumer electronics. What is more amazing is that POSCO is also one of the world’s lowest cost producers of steel. It can achieve this bcos it has the most integrated high capacity steel mill in the world and it also attracts the best talent in Korea to work for the firm. To that end, it even has its own university!

Hence the firm consistently generated free cashflow and paid dividends while having a clean balance sheet with no debt. Having said that, the wheels of fortune might be turning as Hyundai tries to break its monopoly in the Korean steel market while the company had also tried unsuccessfully to expand into the Indian market. In recent times, the dividend has fallen to 2% while free cashflow yield is also below 5%.

So that’s a short summary of one of the oldest industry on earth. In short, it’s best to avoid, as the industry had not been very profitable for shareholders except for the 5 years starting 2003 when the whole world got into a once in 30 year situation whereby there was a shortage of steel. This happens when a big country industrializes after a long drought and no new investment was made in steelmaking. The last country before China was Japan, which started the steel boom in 1970s.

Next on the list is India, but that might be 2030, if we use the once in 30 year rule.

A Girl in the Convertible

There were some academic studies done on capital structure some years ago by two professors. I only remember the study as the M&M theory. M&M being the initials of the two professors. Both professors subsequently won Nobel Prizes! The same theory also talks about dividends, and I thought that the conclusions are worth sharing here.

According to the study, in a perfect world where there are no taxes, no legal or accounting fees and stocks are infinitely divisible, then it doesn’t matter whether stocks pay dividends or not. Bcos investors can just sell part of their holdings whenever they feel like paying themselves some money.

In the bigger scheme of things, it also doesn’t matter what the capital structure of the company looks like. The firm’s capital can be 100% debt or 100% equity or any other makeup, it doesn’t really matter. What matters is that the firm will only be able to generate enough profits to keep it from going bankrupt, and the market price of the firm is always the right price, ie its intrinsic value. And this is the basis of the Efficient Market Hypothesis.

Luckily the world is not like that and dividend matters. A bird in hand is worth two in the bush. Or as Warren Buffett puts it, a girl in the convertible is worth five in the phonebook. So as investors, we want some dividends to come to us, regardless of what Nobel Prize winners theorize.

Hence, I personally like to find good dividend stocks, and hopefully the firm also enjoys a bit of growth over time. The Dividend Aristocrats of the S&P500 is really a good hunting ground for high quality global names. As for Singapore, I have generated some dividend stock lists in the past couple of years. The most popular one is at the right column of the blog.

Some would question why this huge emphasis on dividends? If a high quality firm can compound its growth much faster, it would be wise to let the firm keep the money and use it to grow. This is the excuse most growth co.s don’t give dividends. Even after they become ex-growth, and they happily squander the cash in stupid ventures or M&As.

Perhaps the best positive example is actually Berkshire Hathaway. Since Buffett can compound growth much better than most people, it doesn’t make sense to pay dividends to his other shareholders. However it is difficult to find managers who can efficiently use capital to compound growth better in the first place. So returning excess cash to shareholders or doing share buyback when the stock is cheap is what a good CEO would do.

Paying an ok dividend also signals that the management have shareholders in mind. (This is also called the signalling theory). Ok being like 2-3% dividend yield, which is whatmost of the Dividend Aristocrat stocks are paying currently. The thinking on this would be something like: Well we don’t need ALL the money to grow, bcos we are in such a fabulous business, we can still grow with limited capex and can generate good cashflow too. And so we would pay our shareholders some dividends, while we continue to grow. Just as we did in the last 25 years.

The long and short of this all is that a stock has a good track record of growing its dividend payment is probably one of the best deals out there (if you can grab it at a reasonable price). Which is probably why Warren Buffett holds quite a number of Dividend Aristocrats like J&J, P&G and Becton Dickinson.

From Free Cash Flow to Willingness to Pay Shareholders

A while back, I wrote this post about stocks that have good record of free cash flow and naturally they also gave a lot of dividends. I thought I would just delve a bit deeper into how free cash flow affects dividend while also looking at 1 or 2 other factors.

Basically, to determine whether a company can consistently pay dividends and grow them, we need to know:

1. The firm's balance sheet, esp the size of its debt
2. The firm’s ability to generate good cashflow or even grow it
3. The top management’s willingness in actually paying dividends
4. If it actually gets paid out, what is the yield?

The first thing I usually look at would be the firm’s long term track record in generating free cash flow, which is operating cashflow minus capex, ie the cashflow that is left after deducting money invested in new equipment, new plants etc. The thinking is that the remaining cashflow can be used to pay down debt or pay out dividend. If the firm has no debt (that’s why Criteria 1 is there), then the money should logically flow into dividends.

On the same post, I only managed to screen out 30-40 companies that have consistently generated positive free cashflow. This is out of 700 listed companies in Singapore. This shows how difficult it is to actually produce enough cash to give out dividends.

However, even if the firm can generate cash, if the management is not willing to pay them out, then there is no point talking about it. To check this, we look at the dividend track record. If the firm consistently paid dividends or even increased dividends, then we are going somewhere.

An interesting case in point would be Yeo Hiap Seng (YHS). The free cash flow track record is stellar but it stopped paying dividends since 2006. It looked like they over-invested in the earlier part of this decade and is now using the free cashflow to pay down debt. But what is the management’s stance on dividends? Will they resume it after the debt issue is resolved? Or are they gonna invest into stupid ventures again?

Let’s look at the previous case study: Starhub. The operating cashflow on average is about S$600mn. The capex is about S$200mn. So FCF is close to S$400mn. The firm has been paying around S$300mn in dividend every year. So that’s very good! There is still S$100mn of buffer for it to grow its dividend.

However there is always the lingering debt issue. Starhub has roughly S$600mn of net debt, and equity in the latest quarter is a miserable S$60mn! This explanation on Drizzt’s blog explains that it’s due to some accounting after they merged SCV. The actual equity is S$1bn. So S$600mn of debt is no big deal. Not to mention, if the S$400mn of FCF is used to pay debt, it will take only 1.5 yrs to clear it.

But still it doesn’t make sense. Why does merging with SCV reduce accounting equity by 90%? Was SCV a negative equity entity? Even if there is no actual impact on every day operations, the book value is something that all investors look at. Can something be done to resolve this?

Well, perhaps it would be resolve in time. Meanwhile Starhub gives close to 8% dividend yield.

Which brings us to the last point: the dividend yield. Is it a reasonable yield? For every dollar buying the stock, are you getting enough back in dividends? But the yield is also a function of future growth expectations. A high yield usually means lower growth expectations.

Starhub’s yield of 8% can mean that there is very limited growth left. Or it can mean that investors expect the yield to fall in the future, ie dividend cut. Or it can really be a steal right now, and over time, the stock rises to $5 and makes the yield more reasonable at 4-5%. Which means you will double your money by buying now.

If you ask me, I think the former reasons look more likely. The market is not stupid, bargains like this don’t last long enough for an amateur blogger like me to blog about it. Not to mention that I am definitely not the first one blogging about it.

The best investment, obviously, is a high yield stock that can sustain or even grow its dividend over time. But high yield with growth is an oxymoron. If the firm can grow, it will keep the cash for growth instead of paying it out to shareholders. High dividend yield and growth cannot go hand-in-hand. But for some value investors, the goal might just be to find 1 or 2 of these high yield growth stocks, buy them and hold forever!

Payout Ratio

Payout ratio is simply Dividend Per Share (DPS) divided by Earnings Per Share (EPS), which tells you how much buffer does the company have for it to increase its dividends. E.g. EPS for Singtel is about 22c and DPS is 11c on average for the last few years, so the payout ratio is 50%.

But what is a good payout ratio?

If the payout ratio is 30-40%, which I would say is probably below the global average, then you know the company has some room to increase dividends.

If the payout ratio is close to 100%, ie the company is paying everything that it earned as dividends, then we cannot expect a lot of growth in dividends unless earnings is going to grow significantly.

In Singapore, sadly, a lot of high dividend stocks also have very high payout ratio, ie no further room for dividend to grow unless earnings can grow. But if earnings can grow, then the firm would want the money to invest in growth and not return them to shareholders. Hence high dividend plus strong growth is an oxymoron. These stocks don’t exist. Or they are very rare.

When dividend and payout ratio both gets too high, dividend cut becomes inevitable.

However, cutting dividend is a big deal for many companies in Singapore, bcos a lot of shareholders buy Sg stocks mainly for their dividend and if it is cut, it means disappointment and selling pressure. Of course it also shows that management is not capable of steering the business well enough to pay their shareholders.

In fact, it got so important to the extent that some companies actually issued debt to pay dividends!

The two prominent cases being Singpost and Starhub.

In the early years when Singpost was listed, it was obligated to pay a high dividend yield for some reason that I forgot. So while Singpost’s net profit was just over S$100mn, it has to pay S$300mn in dividends! So bo pian, raise debt to pay dividend. However, that only happened more than 6-7 years ago and the company hasn’t done anything stupid like that ever since.

Ok let’s talk about Starhub.

For the past few years, Starhub generates annual net income of roughly S$300mn and pays roughly $300mn in dividends as well! At the same time, the firm increases its debt holdings almost every alternate year, in 1 year by as much as S$600mn! So people asked questions like why do they raise so much debt when they could have cut the dividends to save money?

Well I guess there are no easy answers. But if we analyse the cashflow statement of Starhub, we can see that perhaps the situation is not as bad. So it might be worthwhile to look at free cash flow vs dividend paid instead. Which is what I would do in the next post!

Singapore's Dividend Aristocrats

The S&P Dividend Aristocrats is a list of stocks in the S&P 500 index that has increased dividends for at least 25 years. As you might guess, this is definitely a very tall order and as of 2010, if I remember correctly, only 10% or 50 stocks made it to the list. It is expected that the list will further dwindle to 40 stocks or so which prompted some to ask whether the criteria is too stringent.

Well, that is the way with our world, I guess, when we don't make the cut, we can always lower the bar ourselves. Haha!

Anyways, some of the names on this prestigious list are very well known, like 3M, Johnson and Johnson, Becton Dickinson, Coca Cola and P&G etc. Thet last four are also famous holdings of our hero.

Intrigued, I went to take a look at which Singapore stocks had a good track record of increasing dividend. Well as our history is not as long, I used a looser criteria of increasing dividend for at least 8 out of the past 10 years.

Here is the list of stocks.

Well, I guess you are as disappointed as I am. Only 8 stocks made the cut. Of which 3 pays a miserable 2% dividend. 2 are in property and construction, a treacherous sector. Another 2 of them belong to the same group entity. And 1 of them is not even a Singapore company.

There are 2 banks, but value investors are very wary of banks bcos there are just too many moving parts to get a good read on these financial beasts. Even Buffett had his fair share of trouble with Salomon Brothers 20 years ago and now Goldman Sachs, which is being criticized for screwing clients left right centre.

Raffles Medical, one of the 2% yield stock, might be interesting but its valuation is way too high at 26x. Might be worth doing some research now and wait for a good entry point.

The truly investable Dividend Aristocrat of Singapore might be SPH but it's main business is in a declining franchise and its high property value is based on frothy valuation in a flood infested shopping district.

In a nutshell, using the dividend arisocrat criteria cannot help us find good stocks in Singapore. We might have some luck elsewhere in Asia.

What constitute an investment?

The way I see it, an investment has to be something that can generate cashflow. Stocks give dividends, bonds give interest and real estate gives rental income. These are real investments.

However in the strange World of Wall Street Craft, anything and everything becomes a feasible investment, an asset class of its own. The recent boom earlier this decade being commodities.

But if you think about it, commodities shouldn't be considered an investment bcos you don't get a cashflow. Holding a ton of copper, or a ton of wheat doesn't give you cashflow. The whole premise is based on prices going up. And when it's based on just prices going up, then it's dangerously close to the idea of the Greater Fool Game. Where you can only make money by selling something that is worth very little, at a higher price, to a greater fool who is willing to buy.

That is why value investors are not interested in price, we are interested in value. Price merely tells us if we can get the asset below its value. If there was no transparent price on the asset, we are happy as long as we have cashflow. But if there is no cashflow, you cannot calculate an intrinsic value of the investment. And in that sense, commodities cannot be classified as an investable asset class. Needless to say, a lot of the newly created asset classes like art, wine, vintage watches and other funny stuff cannot be called investment.

However, if those above mentioned can somehow be construed to generate cashflow, then the story becomes different.

For e.g. if a couple of artworks can be put together at an exhibition hall, and the owner can charge fees for viewers, then we have a cashflow, and the whole business can then be valued. In the same vein, wine is not an investment but the vineyard is. Copper may not be a true investable asset class, but a copper mine or a mining co. is definitely investable.

Similarly, traditional assets that count as investments may not be such if it doesn't generate cashflow. The best examples would be perennial loss making companies. Think Chartered, NOL and the likes.

As the saying goes, cash is king. Well if cash is king, in my opinion, cashflow is then the true master of the universe.

The Story of Bak Kut Teh

Just heads up, this is a post with very little value add. I thought about this while eating Bak Kut Teh. Just for entertainment. :)

For the non-ASEANs reading this, Bak Kut Teh is a Chinese dish that originated in S.E.Asia that is made up of spare pork ribs cooked in traditional Chinese herbs and served in a soup. It is also usually served with rice and soy sauce. You can take a look at my half eaten set below.

There is an interesting urban myth about the origin of the dish. Coolies working in Singapore and Malaysia earlier last century had to do tough work like carrying heavy goods at the ports for the whole day. As they were very poor, eating meat was not an option and eating food lacking nutrition ultimately resulted in poor health, sickness and weak bodies and their ability to work to generate money.

So Bak Kut Teh came to the rescue. The coolies would buy ribs that nobody wanted from the butchers at a cheap price, mixed with simple herbs (also cheap, I think) and then eat with rice and soy sauce. Simple as it is, this dish gives them strength, helps to prevent sickness and allow them to work all year long.

Of course today Bak Kut Teh has evolved into a popular dish that actually costs more than normal hawker food and it is usually served with spare ribs with lots of meat, You Tiao (or fried dough) and other stuff.

So while eating Bak Kut Teh, I thought about investing and realized a few analogies which could be drawn.

1. Value for money

Obviously, value investing is about buying more for less. Buy a good company at a reasonable price. Buy things cheap. Bak Kut Teh stands for this. Well at least during the coolies' times. Cheap but nutritional, helps the coolies stay healthy, get work done, earn more money. Today it is a bit different lah.

2. No Fat

In investing we want to buy companies that are lean and mean with no fat. That's Bak Kut Teh! Companies must understand that they have to stay lean in order to generate good returns. Actually, we ourselves strive to stay healthy and keep those cholesterol away so that we can fend away the harmful effects of metabolic syndrome right?

3. Innovate in changing times

The origin of Bak Kut Teh crystallizes the spirit of innovation. First, the coolies or whoever working with them came up with this spectacular dish that helped improved lives. Then, over the years, the dish itself evolved to fit into society. Even today, it's a highly popular dish. In investing, we must also be on the lookout for companies that can keep re-inventing themselves. As we speak PC and PC related companies are dying, big names like Intel, Dell and Microsoft are going all out to re-invent themselves. Although I must say, it's very hard to predict which companies can successfully evolve.

As with individuals, a lot of people struggle to stay relevant in society bcos things are moving so fast, it just gets harder and harder. Perhaps the story of Bak Kut Teh is really about evolution and how we should always re-invent ourselves to be useful.

High Dividend Stocks

An updated post "High Dividend Stocks 2011" is finally out! Do check it out.

Here is a list of stocks that I generated using the following criteria

1. Market cap more than $50mn
2. Dividend yield more than 4%
3. Past 3 yr average ROE more than 9%
4. Past 3 yr FCF yield more than 7%
5. Past 3 yr EBIT margin more than 4%

Obviously, as with all quant screens, more work needs to be done to refine the results. The first counter definitely look strange. I mean, with dividend of 30%, in 3 yrs you get back your capital. How likely is that?

Chances are something is wrong with the firm or with the data.

The name that I thought might be interesting is Telechoice - which sells prepaid phonecards. With the immigration floodgate opening again, it might see some profit boost in the next few years.

The other safe and well-known dividend names like M1, Starhub, SATS etc, well should always try to buy more of these if you believe in the Singapore story.

Well a small gift to all the teachers reading this blog today!

Valuation Expansion

On Wall Street, a lot of educated monkeys like to talk about valuation expansion. Basically valuation expansion simply means that some stock trading at 15x PE should be trading at 25x PE bcos its industry is sexy, or the company has undergone transformation of its business to become the new growth story or some other cock-and-bull story.

So say the stock price today is $15, and the stock earns an EPS of $1 ie PE is 15x. Valuation expansion simply means that the stock should be $25 bcos PE should be 25x. The basis of this argument is that since the stock is in a growth industry, or has transformed its business, or watever crap reason, the future EPS is not just $1 but much higher. Since we are not sure what that would be, just give it a higher PE to justify this growth.

The ingenuity of this crap theory is that nothing changed, but the "value" of this stock just expanded 60%. This then can be used to justify buying the stock at any price bcos we can always assuming super normal growth and increase the valuation. We can even increase the target multiple further from 25x to 50x. This would expand the original "value" by 333%.

Let's just do a simple experiment the debunk this valuation expansion theory.

Yr 0 EPS $0.5 (Stock price $25, ie PE 50x)
Yr 1 EPS $1 (Stock price $25, ie PE 25x)
Yr 2 EPS $2
Yr 3 EPS $3
Yr 4 EPS $4
Average EPS $2.1
True intrinsic value (using PE 15x) = $2.1 x 15 = $31.5
Upside = $31.5/$25 = 26%
Upside per yr roughly 5%

So assuming today we are at Yr 0 and this company started out with an EPS of 50c but bcos of its super power growth, the stock market has already valued it at 50x PE current yr and 25x next yr. Of course this is assuming it didn't disappoint, its EPS doubled to $1. In fact it didn't disappoint for the next 4 yrs and its EPS grew from the initial 50c to $4. This stellar firm actually grew its earnings 8 folds in 5 yrs!

Now how spectacular can a normal company get? I would think this type of growth puts the world's best growth firm to shame. Look at APPL, the darling-est growth stock in our lives. Currently it's the 2nd biggest company in the world by market cap. Its net profit was $1.3bn 5 yrs ago. Last year, it was $8.3bn. The 5 yr growth alas is 6.4x, still a tad less than our hypothetical co. at 8x.

So 8 fold increase in EPS is really as good as it gets. But, if you have bought it for 25x or more, the return in stock price is likely to be single digit return. As I calculated, the intrinsic value is $31.5 vs today's price at $25. Of course, the stock might bounce up a lot, to say $50 then collapse, or it's price might continue to stay very much higher than its true value of $31.5, but we are value investors remember? We don't follow prices. We follow value.

What I am trying to say is this: when you buy a valuation expansion story, your rate of return is destined to be meager even if the story comes true. In our hypothetical case, the stock return over 5 yrs is about 5% per yr. How fantastic!

To me, valuation expansion is then just another variation of the Greater Fool Game. Valuation expansion means the earnings of the company is not great now, BUT bcos there are a lot of people willing to pay 25x now, therefore the stock price should rise by a huge magnitude.

I would think that a better way to make money would be the always buy below PE of 17-18x. Value investors would certainly do that. With valuation expansion, there is no margin of safety at all. What if the growth didnt come true? What if the genius CEO died?

So when you hear valuation expansion next time, pls beware!

PS: APPL did trade at 25x PE 5 yrs ago and you would have made a lot of money buying it at 25x 1 yr fwd and held it until today. But the better chance to buy was during the Lehman Crisis when the PE fell to 15x 1 yr fwd and you could have more than doubled your money in 2 yrs!

Did Buffett underpay Mrs B?

I did a post some time back highlighting that perhaps the acme of investing might actually be not to short-change anyone in any transaction. ie to buy a stock at its intrinsic value while waiting for the intrinsic value to grow. This is a very sensitive point when we are dealing with day-to-day business owners rather than the stock market.

Obviously buying at intrinsic value is also different from buying a company at a huge discount and then waiting for it to revert back to its intrinsic value, which is the original thesis of value investing.

I thought that Buffett might have just pulled this off, and that is why he is the greatest investor on Earth.

Recently I did some study on the exact transaction that he did: buying over Nebraska Furniture Mart from its long-time founder and operator Mrs B.

The story goes something like this. In 1983, Buffett on his birthday, simply went in to the Mart and asked Mrs B. if she would like to sell, and at what price?

Here are some no.s at the time the transaction was done.

NFM
Sales 100mn (reported)
NP 5-10mn (estimate by 8%pa)

GPM of NFM 60% (reported)
Furnture mkt OPM 10% (industry average)

$55mn of 80% stake (reported)
$69mn for 100% stake (reported)

Buffett paid 7-14x PE

Based on my estimate, Buffett probably paid 7-14x PE for NFM. Now this is a huge range. If he paid 7x, he would have obviously underpaid Mrs B. If he paid 14x, then it's probably fair. So did he underpay Mrs B?

Well, sadly we will never know for sure, but chances are Buffett probably did underpay somewhat, judging by his track record and considering how the firm grew over the past 30 years. However we must also note that Mrs B was a willing seller at $69mn. She quoted Buffett that price.

Also, Mrs B probably would not have gotten much more if she were to list her company in the stock market. The brokers will take a cut. She has to pay for some auditors. Perhaps hire a few more staff to handle Wall Street people etc. With Buffett, she got her $55mn check on that day. And the best part, she continues to do what she does bcos Buffett wanted her to continue to run the business. No Wall Street analysts on her back every quarter!

I guess the conclusion here is that Buffett might not have been as noble as I thought. He did underpay his acquaintance somewhat but judging from the circumstances, Mrs B didn't mind that she got a couple of millions less (that is assuming she actually bothered to go and do a thorough valuation of her own co.) and she gets to do her job exactly the same way she liked it.

In fact, a couple of years later, Mrs B sold another business to Buffett. If she felt cheated, would she had gone back to Buffett?

The real lesson learnt for me here would actually be thinking about the best solution for both parties in every transaction. This means disclosing all the pros and cons to each party about the transaction, not witholding any information at all and working out the seek the win-win situation. I think this is very possible if both parties are rational, honest and committed to a long term relationship.