The vagaries of value (Amended)

EVERYBODY claims to practise it. Certain fund managers swear by it.

Numerous websites and trainers claim to teach it, some charging a fortune for imparting their knowledge.

But what exactly is value investing? And how can the average investor practise it?

Today, we kick off a discussion on these two probably most misunderstood words in investing jargon.

Value investing is a topic that cannot be covered in one page, and we will discuss it in the coming weeks and months.

For this week's column, we examine value investing from our first source, which is of course Benjamin Graham and David Dodd's 1934 textbook, Security Analysis.

The text has inspired numerous followers. The most famous is Warren Buffett, who studied with Graham, worked for him, and applied his principles to become one of the most successful investors in the world.

One main idea of value investing is that every security has an intrinsic value.

Think of intrinsic value as an anchor in stormy seas.

Sometimes, the market offers the investor an opportunity to buy a security at a substantial discount to its intrinsic value. The savvy investor will accumulate the security.

At other times, the market will offer securities for sale at prices significantly above their intrinsic value. The savvy investor can then unload his holdings for a significant profit.

That sounds so simple.

But determining intrinsic value, as Graham and Dodd pointed out in the very first chapter of the book (see sidebar), is easier said than done.

Intrinsic value is not book value

The first thing they point out is how intrinsic value is not the book value of a company.

"Neither the average earnings nor the average market price evinced any tendency to be governed by the book value," Graham and Dodd wrote.

Book value here refers to the accounting equity value of a company after its liabilities are subtracted from its assets. What is due to non-controlling interests is also taken out.

Another name for book value is net asset value, or total equity attributable to owners of the company.

Are Graham and Dodd right, in that book value is a poor gauge of intrinsic value?

Many of us continue to assume that a company's net asset value per share provides some kind of guide to the value a company should trade at.

For example, we use book value to evaluate property companies, banks, and traders.

To be clear, investors don't necessarily think that the fair value for a company is exactly one time book.

But they do think book value is a good guide, and this can be a problem. Just look at how banks are valued, for starters.

In a report on Jan 3 downgrading DBS Group to "hold", OCBC Investment Research wrote that for DBS, its five-year average price-to-book (P/B) ratio is around 1.15. This means the stock has tended to trade at a 15 per cent premium to the net assets recorded on its books.

"But with a still mixed outlook for the region, we think that for the near term, the stock will continue to trade at a discount to this level, likely within the 0.95 time to 1.05 time P/B band," the report said.

"With the recent re-rating of banking stocks, we have moved our valuation peg to 1 time book, resulting in a fair value estimate of S$17.83.

"At current levels, we are downgrading the stock to a 'hold' and think a better level to re-enter the stock will be at S$16.80 or lower."

In this example, the value of DBS is pegged upon at least three premises:

1) The five-year historical book value it has traded at;

2) Current circumstances that warrant a discount to that level; and

3) The analyst's estimate of where DBS's book value will be at end-2017.

Taking the first premise, we can already see how it is debatable whether history is a good guide to the present or the future.

DBS has traded at a far higher valuation of 1.5 times book value or higher during 2006-7, when Asian growth prospects seemed brighter. In fact, its 15-year average P/B ratio is around 1.25 times.

Picking the right historical period to use as a benchmark is vital. After the global financial crisis of 2008-9, the global economy has slowed dramatically along with trade. Populations continue to age, and there is a general consensus that growth will be slow in the years ahead.

If a paradigm shift has indeed occurred in the form of a structural slowdown, it will not make sense to use historical measures of average value that stretch too far back in time.

On the second premise, we see how OCBC argued that an uncertain outlook for the region means a discount needs to be applied to a firm's valuation.

But this requires two judgement calls: Whether you think the uncertain outlook will last long enough for a discount to be warranted, and how big a discount needs to be applied.

On the third premise, it bears noting that the "1 time book" valuation of S$17.83 in the report is the house's estimate of where book value will be at the end of the year, instead of the last reported book value, which was S$16.68 at end-September 2016.

Therein lies another judgement call.

Shifting book values

More fundamentally, what is DBS's book value here?

DBS is a big bank. Its assets are mostly the loans it issues to businesses and consumers, from which it gains income by charging an interest rate.

Its liabilities are mostly the deposits that consumers and institutions keep with it, on which it has to pay an interest rate to keep customers interested in keeping their money with the bank.

While the bank's liabilities are absolute, meaning that it has to honour the obligations to avoid going bust, the value of its assets is less clear cut.

For example, the bank makes thousands upon thousands of loans, and it is impossible to tell whether every single one is sound.

In an economic downturn, it is impossible for an investor to know for sure exactly what proportion of DBS's assets will get impaired.

We have to guess, for every loan, whether the customer can pay the bank back. And if it can't, we have to guess what price the bank will get after seizing the customer's assets and selling them. In other cases, the bank will work with the customer on a repayment plan.

Net assets can drop suddenly. In the Swiber debacle, when the oil services firm collapsed last year, DBS revealed its total exposure to the firm was over S$700 million. It had to take a S$150 million provision for losses related to the loan.

Loan loss provisions will reduce net assets. On the flip side, if the economy is booming, banks can create new loans and increase their interest income, boosting net assets.

With assets and liabilities constantly changing, we see how book value is, to put it mildly, an extremely imprecise measure of how much a bank is worth.

Yet in the case of banks, book value is still a useful starting point.

Investors just have to make plenty of judgement calls and modifications to come up with a valuation.

We use history for comparison. We compare to peers. We compare net interest margins - the spread between what a bank makes on its loans and what it charges on deposits - and ascribe higher valuations to more profitable banks.

We say some banks should be valued more because they are exposed to regions of higher growth, or because its products have gained more traction among consumers.

Some, like Germany's Deutsche Bank, might be so complex or mired in so many problems that we think a 60 to 70 per cent discount to book value is warranted.

Book value is highly problematic, yet investors use it anyway.

A starting point

So what is intrinsic value?

Some form of the company's potential earnings in the future will also have to come into the picture, along with its assets.

Graham and Dodd provide an example.

"In 1922, prior to the boom in aviation securities, Wright Aeronautical Corporation stock was selling on the New York Stock Exchange at only $8, although it was paying a $1 dividend, had for some time been earning over $2 a share, and showed more than $8 per share in cash assets in the treasury. In this case analysis would readily have established that the intrinsic value of the issue was substantially above the market price," they wrote.

Wright was trading at under four times its earnings. A casual observer could have noted how even if the company does not pay its owners all of its earnings, it just needs to keep up its tiny dividend payment for eight years for an investor to make back his money.

Assuming there are no significant liabilities, and that earnings do not collapse, it does seem like Wright is dramatically undervalued. We just don't know by how much.

Let's see what happens next.

"Again, consider the same issue in 1928 when it had advanced to $280 per share. It was then earning at the rate of $8 per share, as against $3.77 in 1927. The dividend rate was $2; the net-asset value was less than $50 per share. A study of this picture must have shown conclusively that the market price represented for the most part the capitalisation of entirely conjectural future prospects - in other words, that the intrinsic value was far less than the market quotation."

Here, Wright is now valued at 35 times its past year's earnings. It will take 140 years for an investor to get back his money through dividends.

It is thus fair to conclude that investors are placing plenty of hope in the company, and the company might be overvalued if its business cannot continue growing.

So while we see there are tremendous difficulties pinning down an exact value of a company, there are nevertheless situations where a judgement call can be applied.

The Wright example might be an exaggeration. But similar situations have occurred in financial markets and are likely to occur again.

In coming columns on value investing, we apply this understanding to securities listed on the Singapore Exchange.

Clarification: The discussion on how bank balance sheets are affected by bad loans and new loans has been amended for clarity. The article above has been revised to reflect this.