Reading financial statements (part 2): Watch out for the Small Print

Some people only read the front pages of the annual report. They're mad. (Either that, or in my experience of working in investment, they're traders.)

With a few honourable exceptions such as Unicorn (when Peter Webb was there), and TR Property, most companies emit annual reports with relentlessly upbeat spin - and the only way you will find the truth is to go straight to the accounts. Even then, the accountants have polished up the statements - it's the notes to accounts where you'll find all the bodies buried! It's also there that you find some of the more interesting statistics to use in analysis.

For instance, the number of employees and the aggregate salary is always noted. While that's not always terribly useful, in a labor-intensive sector it can be quite interesting. It also gives you a good handle on whether the directors have actually achieved the cuts they told you about last year - I know one company that has been 'removing duplication' between its UK and Chinese operations for years, but the number of employees has steadily increased.

An area that is being studied increasingly closely is the note on borrowings. Right now it's covenants and other limitations that are probably the focus of attention; however if interest rates rise, you'll also be able to check up which companies have fixed debt and which are exposed through floating rate debt.

Cash is a fairly simple concept, you might think. But be careful; not all cash is available funds. Here again, trawling through the notes you'll find out whether any of the cash is already spoken for - for instance, client money that is held in segregated accounts, which appears on Ig Group's balance sheet. IG Group appears at first sight to have £520m cash, but once you've excluded client money the figure comes down to £99m.

You also need to pay close attention to the notes when it comes to capitalisation of expenses - this is where you really need to clamber over the fixed assets and the cash flow account to see how the capital expenditure relates to the additions to fixed assets. I've come across a number of media companies that underdeclared their fixed running cost because they capitalised concept development costs - that's fine as long as, say, one in five of your concepts is commissioned and goes into production, but if suddenly you pitch ten without any takers, it can catch up with you. During the internet boom, quite a few ISPs decided to capitalise all their marketing costs - again, once growth disappeared, the consequences of that decision came back to bite them.

I always try to even out the effects of capitalised costs to come up with an 'as if' number showing the actual cash costs of running the business. It's surprising how often it's significantly different from the declared cost.

Watch out for depreciation policies, too. It's tempting to flick through the notes on policies, because they are usually boring, and the same as last time. But depreciation is one of the easiest numbers to fiddle about with; it's also something that can make a big difference to the apparent profitability of asset-intensive businesses such as equipment rentals or  datacentres.

For instance if you have a datacentre, and you assume the useful life of your kit is 5 years, for every £1000 of kit you'll be writing down £200 each year (if you're doing so on a straight-line basis). If, on the other hand, you say it's going to last 10 years, your costs will be £100 a year lighter - which could mean your results look 10-20% better than your rival's. I suspect that while some of the trunking and cabling might last ten years, the hardware won't - which means those results are going to look atypically good, and don't represent the real state of play.

Equally, companies occasionally change the depreciation policy and though I'm sure they don't do so just because it will make this year's EPS look better, that comes as a welcome side effect. Even something as simple as changing from a straight line basis to a reducing balance basis can have quite an impact on profits.

Apart from the accounting treatment used, notes also tell you a number of interesting facts about the way the company has been operating. Two I've found particularly interesting and, on occasion, profitable are the notes on related company transactions, and on contingent liabilities.

The note on related company transactions is usually quite late on, about note number 18 or 19, and it's usually anodyne. That's good - this is one of those times that boring is good, and it's boring 99 percent of the time. But just occasionally it's interesting - and that's usually a bad sign. For instance software company KMS seemed to have increased its revenues quite handily - until you looked at the related company item. It had bought a stake in a company called, I believe, Custard Cream, and the company had used the funds to buy KMS software. In other words, it was giving customers money to buy its product - not a million miles off handing it out for free, except for the fact that it could still book revenue. I really didn't like that, and I was not surprised when the company failed to thrive.

Another company which had a habit of helping its customers to afford its services was Guinness Peat Aviation, which came to the market with a big splash only to implode not long after.

On the other hand sometimes you see a director's other company has sold a few boxes of biscuits to the company canteen... or the equivalent. If it's a small item, I would usually ignore it, at least in a smaller company where occasionally taking a helping hand from another business can be cost-effective (in terms for instance of sharing secretarial and book-keeping costs).

Accountants are conservative little creatures, and if they can foresee a loss, and work out how much it will be, they will write it off against the P&L. But even so, there are some potential losses that won't be written off - but might be included in a note on contingent liabilities.

These are liabilities that probably aren't going to crystallise. For instance, a company might sell a business and guarantee the level of profits will not undershoot a certain level in the next three years. It may set that level 50% below what the business was already making, and think there's no way it could perform worse than that - but it should still show a continent liability, just in case.

Now of course in times like these, some of those liabilities that were thought improbable a few years ago are no longer improbable. They're likely - some of them are happening. That's where if you read the notes to the accounts you'll be months ahead of investors who only worry about these liabilities when they hit the headlines.  A software company that has a money-back guarantee for customers might not expect many of them to activate the money-back clause - and when times are good, clients might not do so. Bring on a recession, and if enough clients do decide to do without the software, then that contingent liability might crystallise.

Unfortunately, there is a huge cost to gaining these investment insights. The notes to the accounts are probably the most boring reading ever devised. If you suffer from insomnia, this will help. It's tough being an investor. But actually, it's even tougher losing money, which is why you really ought to read the notes.