In a market where the forces of value and momentum are moving to historic extremes, it is fascinating to watch the bulls and bears on the bank sector argue their corners. Not surprisingly, the bull corner is less crowded than it was one or two years ago. The lexicon of a bank analyst has exploded in the last six months alone and a research note is incomplete without mention of SIVs, conduits, collateralized loan obligations, collateralized bond obligations, Alt A mortgages and so on.

Analysts who were previously bulls and who had highlighted low price earnings ratios and high dividend yields as attractive are now spooked by the write offs and impairments the banks have had to take on these esoteric assets which they hold. In fact, it is quite easy to paint an apocalyptic view of the future with further write-offs eating into the balance sheets of the banks to such an extent that their long term survival would be in doubt.

We are often asked questions such as: have we seen all the write-offs? How much worse could it get? Are these write-offs fully discounted in bank valuations yet? The only answer to all those questions is that no-one knows. Write-offs (and their half-sister impairments) are based on a large number of assumptions and these assumptions can produce massive differences.

Imagine a pool of sub-prime mortgages. To calculate a potential write-off, one needs to assume how many customers are likely to move into arrears, how many of these will ultimately default - and therefore see their house repossessed -  and how much the houses will be sold for once repossessed. An auditor can check the assumptions used in the model and take a view as to whether they are ‘sensible’. He may possibly even be able to take a view on whether they are prudent. However, he has no ability to tell if they are correct.

One possible short-cut is for banks and their auditors to take their lead from market valuations of assets similar to those which they are trying to value. While this may appear an attractive approach, many commentators would claim that market prices – which are usually based on indices – are being affected by both panic selling of similar assets and the use of the indices as a hedging instrument (as nervous holders attempt to hedge their overall exposure).

It is no longer just sub-prime mortgages, and the derivative products borne from them, whose value is in doubt. Better quality residential mortgages, commercial mortgages, investment grade bonds, high yield bonds and leveraged loans have also been wrapped up, repackaged as bonds and sold to a wide variety of investors. And all components of all of these packages are falling in value.

It is therefore not surprising that analysts formerly bullish have now changed their tune. They highlight ‘lack of visibility’, ‘lack of faith in management’ and ‘lack of potential catalysts’ as reasons for their new-found bearishness. The implication is that any sensible investor should wait for improved news and a line to be drawn under the write-offs before dipping one’s toe into the sector.

While this approach sounds intuitively appealing, it is one which is of less value in the real world. Only history will tell us when the write-offs ended. Obviously management could try and call the bottom at some point but surely investors would be unwilling to take their word – after all, many feel they have been badly misled by similar words in the past. The stock price bottom is by definition reached at the point of maximum bearishness. But once again this is something that can only be measured with hindsight. Instead, we believe it is vital to look for clues that we are there or thereabouts as history tells us that huge price increases are possible before any good news becomes apparent.

What clues are we looking for? With banks earnings so volatile and their balance sheets under such pressure, we believe that short-term earnings and dividend measures are of very limited use. Our preferred alternative is to compare the share prices of the banks with their net assets. Historically, banks have traded at multiples of their asset value. This recognized their high levels of profitability which itself reflected the oligopolistic nature of the sector.

Some banks such as Bradford and Bingley have traded below their asset value while others such as HBOS are fast moving towards theirs. However, these valuations are obviously low for a reason; investors have very little confidence in the asset values. This is hardly surprising given the highly geared nature of banks’ balance sheets.

For example, HBOS has, according to Deutsche Bank, assets of over £500bn on its balance sheet compared to approximately £18bn of tangible equity. Clearly a little stressing of a spreadsheet can generate some scary write-offs which can quickly eat into the tangible equity.

We are therefore left in a position where we (and no-one else) can speak with sufficient certainty to be confident that the banks balance sheets are big enough to take everything that is currently being thrown at them. Perhaps, the write-offs will bring one, or even a number of banks around the world, down. This is possible but is clearly a worst-case scenario. And if followed to its natural conclusion anyone looking to invest with greater certainty should never invest in a bank.

However, we are not that apocalyptic. We do though believe it is important to be selective and hold just three banks : HSBC – for its balance sheet strength; HBOS, as we believe many of the market’s fears over its ability to fund its loans are overdone and Lloyds TSB as we believe the weakness of many of its competitors in the mortgage market will allow it to significantly increase its market share.

This article is not investment advice or a recommendation to buy any share or fund mentioned directly or indirectly, and should not be relied on in making any investment decision.

Please note that the opinions expressed in the above article are the opinions of the author alone, and do not necessarily represent the views of Investec Asset Management or its associated companies.

Investec Asset Management, or any of its associated companies or employees (including the author), may hold positions in any of the securities mentioned in the above article or in related securities including Temple Bar.

For more information on Temple Bar visit www.templebarinvestments.com