Since the yield gap between gilts and shares has reversed I wonder what investors use to evaluate shares. For example I used undated gilts as my discount rate and projected dividend growth for 'x' years to arrive at a fair present value. I do not use PEG as it is mathematically naive and does not explain zero growth or potential dividend cuts.
Growth shares make most sense but what rating? One can have a multiple of earnings or dividend but at what rating would you cap a 10% growth stock? A crude rule of thumb was to add 10 points to the PD ratio. Thus if undated gilt is 25x dividend the 10% growth stock should be worth 35x its dividend. The conundrum also is that if you figure 10% growth perpetually then the share price approaches infinity.

Let's compare two house builders Telford Homes and Persimmon, we will use the pe ratio, housebuilders are valued on net assets also so no one should be encouraged to buy from the following research.

If we compare the current pe of both companies

PSN 12.31
TEF 26.06

both shares have a good run lately and comparing the figures u would think PSN offers the better value and wonder why the market has bid TEF so high. U may even think that the market might bid PSN to the same pe which would mean the price could double.

The pe's shown are historical, based on the last published accounts. To see what the market thinks may happen in the future we need to look into the fcasted pe's.

PSN

2010 actual 42p eps > pe 12
2011 fcast 34p eps > pe 15
2012 fcast 42p eps > pe 12

By looking at the fcast pe we can tell that the market expects PSN to make less profit this financial year than last and looking at the fcst eps this confirms that.

PSN has traded in the past on a comfortable pe of around 15, so if we look into later this year, the market may value PSN 42p eps x pe 15 = 630p
May be more as they have traded on a higher pe or less because they have traded on a lower pe.
One thing to consider, when they reached their recent high it meant everybody who had bought since June 2008 was in profit so profit taking was to be expected sooner than later.

Thanks Charttrader. I do know about the crux of your example. My thought was what someone such as yourself would pay or what criteria you would use if most evidence suggested that a share would most likely grow by 10% per annum for 5 years.Thus current eps 10p but pretty sure it will go 11p, 12.1p, 13.2p, 14.4p and 15.8p. A super confident guy may pay 10x year 5 at 158p, someone else pay not be confident and go for 110p. The market is a mix of these types and the price may bounce between these levels. If 10% could be had in perpetuity then 158p is a bargain. Tesco is a good example of 10% growth and the price plummeted because holders of the stock hauled in the horizon for growth.
Likewise if one can find it how would you price a 5 year 20% grower versus a 10% grower? At what level would you prefer the 10% er over the 20% er? Say my 20% share goes 10p, 12p, 14.4p, 17.2p, 20.6p, 24.7p. The key is also to maintain a rating, and the higher the rating the more difficult it is to maintain it.
By the way out of builders group Bovis is perplexingly overvalued and Taylor Wimpey looks cheapest But all take good recovered results as a done deal,; they certainly have improved margins and ASP. At least Persimmon pays a div but they have maxed at 500p+.

I don't look at gilt yields because it's one more variable over which I have no control or knowledge of what it's going to do in the future. Discount rates are also dangerous in many cases because you can create whatever valuation you like by fiddling with it.

On a more positive note I agree with your basic premise that stock valuation is mostly about dividend growth and the underlying earnings which support it; although I will say that there are a million and one ways to value a stock, but dividend growth is one of the big ones.

You mention that you're talking about a stock where you're pretty sure the growth will continue, like Tesco for example, so I guess we're talking about a selection of stocks that have a proven history of growth (again like Tesco?). I think that's a pretty sensible starting point.

One way to do this is to use 10 year data - so you'd look at 10 year earnings and dividend growth and how stable and progressive they've been.

Once you find those kind of stocks the next question is the question you started with - Exactly how do you value them?

Again there are a million and one ways to do it, but a couple that I like are:

PEGY10, which is basically PE10 / (G10 + Y), where PE10 is the 10 year PE ratio, G10 is the 10 year growth rate and Y is the current yield.

I think that's a reasonable starting point, but of course it doesn't create a £ value of the stock but rather a ratio to compare one to another (or perhaps to compare against the FTSE 100 index to make sure it's worth taking the extra risk of investing in a single company).

Another approach that I like which is quite a bit more complex is to rank those 3 factors for the whole universe of stocks and then combine the ranks which gives a similar but more stable comparator than PEGY10. But like I say this is a bit more complex and time consuming to put together. You then just exclude all the stuff you're not interested in until all that's left is the quality dividend growers sorted by this multi factor rank.

Or more simplistically, just compare long term growth, long term PE and current yields between different companies that you're interested in and also compare them against the index.

Hope that's of some use.