The railways led to the standardisation of time across the UK. Prior to that every town and city had its own timezone. There was a story I once heard that went like this:

‘Engineers gradually built a telegraph network across the UK to link up stations on the rail network. As well as allowing communication, the telegraph allowed stations along the railways to synchronise their station clocks. However when the engineers arrived at one small village they found that the station clock was already at the correct time. Naturally curious how this could be they asked the station master his secret. He explained that every morning he stood on a chair and looked across the road through a window above the front door, into the hallway of the house opposite where Mrs Smith had an expensive grandfather clock. He then simply matched the station clock’s time to that of the grandfather clock. “It’s expensive so it must be right”, he said. Naturally curious the engineers rushed across the street to ask Mrs Smith how her grandfather clock was so accurate. “Very simple”, she said. Every afternoon she opened her front door, looked across the street and set her grandfather clock to match the station clock. She pointed out that “Everyone knows that station clocks are accurate.”'

This amusing anecdote illustrates a key aspect of valuation that is relevant to markets. Many prices are based on nothing more than observing other prices and having confidence in those observations.

Mrs Smith and the station master had no agenda other than time keeping so there was no bias to alter their time keeping in one direction or the other. In markets there are often implicit biases in one direction for all market participants.

Price discovery

The chart (right) shows a carefully selected 14-year window for the IPD UK (Commercial) Property Monthly Total Return series and a ‘mystery’ time series.

Let us consider the property index first. It is typical when looking to sell a property (or indeed to rent it) to obtain estimates of achievable price from three agents. Each agent will in turn consider the price at which other properties in the area have sold (or rented).

Naturally every seller is keen on maximising the price he achieves for his property when selling - and will choose the agent who suggests the better prices. An agent who is wildly bullish is likely to be excluded from the selection so over time agents who want to get ahead will carefully propose prices that are a little, but not too much, above the prevailing ones in that local market.

Many buyers, curiously, have more confidence in buying in a market that is gently rising rather than one that is falling or too hot.

There are other behavioural aspects that can also impact this example - in particular sellers, unless distressed, do not have to sell unless they achieve the price they seek.

In each case prices are referenced to ones that have been transacted in the locality. This is no different to Mrs Smith and the station master, except that this time all parties (agents, sellers and buyers) have a vested interest in prices rising.

Lessons to learn

Image that you were a lending officer at a major bank during the period shown in the chart. You were in essence lending against rising prices. There was never ever a down month (on a total return basis - remember that I carefully chose the index and the time window), volatility was low and the increase in prices was steady.

It is very easy to understand how, in such an environment, lending standards would steadily slip, and move from say 60% equity, to 50%, 40% and eventually less than 10%. The consequences of that are well known, with the property crash that occurred and the impact on the banking sector.

A very common valuation technique that all investors need to be familiar with is ‘relative valuations' and the comparison of the price of a stock against that of its peers. Typical valuation tables can be based on price to earnings, price to book, dividend yield, enterprise value to earnings before interest taxes depreciation and amortisation (EV/EBITDA) or a myriad of other parameters.

Just like Mrs Smith and the station master these tables assume that the other data points are accurate. Naturally brokers and others who are keen to market or sell a stock will, like the property participants we discussed earlier, select members of a valuation table to show the stock or the investment they propose to be cheap in comparison to the rest of the table.

This does not mean investors entirely throw away relative valuation tables (and focus solely on absolute valuation) but they must be careful to understand how valuations are derived and cross-reference them to absolute values. For instance if a stock is 10% cheap to the market but the market is on 30 times earnings then actually the stock is probably not cheap at all.

Mystery resolved

The mystery series is Fairfield Sentry Ltd, a feeder fund into convicted fraudster Bernard Madoff’s vehicles. To put this another way - if Bernie had invested in UK commercial property he could have had better returns than he did, and they would have been entirely legitimate. The correlation between the two time series shown is 96.7% and to the end of 2003 it had been 99.7%. Some might even suggest this is even more curious than the case of Mrs Smith and the station master.

By Randeep Grewal

After 14 years in the financial markets as an analyst and portfolio manager in London working for a top hedge fund and then the world's oldest fund management company, Randeep Grewal is now involved in a number of earlier stage companies as an advirsor, mentor or investor. contact@grewal.com