Trailing, TTM and forward figures e.g. PER, EPS, etc.

[TheMotorcycleBoy]

So as many of you know, Mel and I have been striving to pick "decent" equity investments by reading company reports and taking account of key figures and ratios. Yes it's a steep learning curve and I don't deny it will probably take many years to ascend.

What I've been trying to get my head round of late, is one of trailing and forward figures, primarily regarding PE, although presumably they could apply to other figures too (e.g. EPS). In addition to figuring out their calculation, I'm also curious of the relative perceived merits of either measure.

I'm now going assume that when people (i.e. publishers of financial data and analysis) refer to "trailing PE", they actually mean "TTM PE", typically, where TTM means trailing twelve months, or to simple folk, the "last year mate".

I'm now going to look at an example, Marshalls Plc (MSLH) a construction material manufacturer and figure out their TTM PE, feel free to comment, criticise etc.

So the price bit is easy: I just lobbed in "marshalls share price" in Google search, and hey presto, I'm given a price of 440 GBp. Ok, so P/E is price divided by earnings per share right? And for arguments sake we'll say earnings means profit after tax to equity shareholders (not non-controlling interests), and that the number of shares is "the number of shares accounting for dilutive effects". I believe that the dilutive effects bit is in order to add to the number of regularly traded shares, those shares which are also on the market, particularly due to a lot of companies these days giving their employees share options, i.e. free (or very low priced shares) with the option to hold or sell etc.

So for the number of shares, I decide to use the figure "last printed", well the last number I could find. That being the "full annual report" published in December of 2017 i.e.

https://www.marshalls.co.uk/documents/r ... report.pdf

I just search for the word "diluted" and ta da on page 92, "Weighted average number of Ordinary Shares (diluted) 2017 198,902,816". Actually I've just found that browsing MSLHs investor pages, has bought us a more up-to-date share count of about 199.4M.

But how do we deduce the "earnings value"? The report I linked above on page 72 (and page 92) informs us that the earnings figure we desire is 42.5M GBP.

But since this figure was valid in the "year ended 31 December 2017" (see the top of page 72), then how we do account for the last 9 months? Since presumably this is what we require to account for a "true" TTM?

So looking further at: https://www.marshalls.co.uk/investor/investor-centre A ha! To the right of the screen I see that on August 2018 an interim announcement was made and reading the PDF, we see that we half year results (12/17-06/18).

I see a figure of 32.5M GBP profit before tax for 6 months. Lets x2 to derive a 65M GBP annual hypothetical value, i.e. assume no change in growth rate etc. And noting on page 27 of the 2017 report "The effective tax rate was 19.1 per cent", I'll calculate that this 65M for a year would be

65 * 0.81 = 52.65 after tax

and looking again at the 2017 it looks as if the non-controlling interests have about 1% stake in the earnings, so I'll leave the 52.65 as is.

So how to make a reported 42.5M for the year ended 2017, and a hypothetical 52.65M for year ended 2018 into the TTM which would be needed for a TTM PE calculation? Where I'll assume 3 months comes from the 42.5M reported up to 2017, and 9 months from this part reported/part guessed year of 52.65M.

So

(42.5M * 3/12) + (52.65M * 9/12) = 50.13M

Is this the kind of process that stock analysts would use? I can't see many other ways, other than guess work, or extrapolating one or more prior reports' data?

So finally my estimate for MSLH's TTM PE is:

4.40 / (50.13 / 199.4) = 17.05

Make sense?

Matt (and Mel)

PS yes, we do hold MSLH shares

[tjh290633]

Don't forget that there may be several versions of earnings in the same report. Which do you use? Adjusted, statutory, or what?

If you were a dedicated follower of a share, you would have the recent quarterly or half-yearly reports to hand, so you can take the ones covering 12 months and add up the earnings per share in each, in order to arrive at the figures for the trailing 12 months.

Since they usually give earnings per share, it seems a little daft to go back to first principles to establish that figure.

TJH

[TheMotorcycleBoy]

Don't forget that there may be several versions of earnings in the same report. Which do you use? Adjusted, statutory, or what?

If you were a dedicated follower of a share, you would have the recent quarterly or half-yearly reports to hand, so you can take the ones covering 12 months and add up the earnings per share in each, in order to arrive at the figures for the trailing 12 months.

Hi TJH,

The marshalls report only seems to list one "version" of earnings in the consolidated income sheet. Though I am aware of the different listings of earnings see here where Sage and Dairy Crest, confusingly paint pictures from varying angles.

We could subscribe to companies of interest for reports emailed to us of interim reports. Do you speculate that's what the analysts?

Since they usually give earnings per share, it seems a little daft to go back to first principles to establish that figure.

Ha ha!

I could ask you "what's your point?" But I won't. Suffice it to say, I'm genuinely intrigued by these datas, what they mean, and the accuracy of their derivation. That's enough for me.

So tell me, in your opinion, does the algorithm I've iterated above seem accurate to you?

[SalvorHardin]

Your method is fine. TTM refers to the past twelve months. It's easier to calculate TTM eps for American companies because they produce quarterly reports, rather than half-yearly. It's much easier to forecast full year eps when you've got published accounts for the first three quarters!

But watch out for seasonal effects; all too often these have a major effect on eps for an accounting period, so assuming that H2 will be the same as H1 can be a faulty assumption (e.g. ice cream sales in winter are much lower than in summer). You can often spot seasonal effects if you compare with previous years' profit and loss accounts.

An example of company which has major seasonal effects is the railroad Canadian Pacific, where the harsh winter weather in Western Canada often has a major effect on Q1 figures (lower sales plus higher operating costs equals much lower eps than Q2, Q3 and Q4 when there's much better weather (and harvested grain to transport)).

Beware of extraordinary items in the eps. These generally represent one-off gains and/or losses on things like asset sales and goodwill write-offs, which won't be repeated next year. Some companies love to trumpet eps including exceptionals if they are positive, but when there's an exceptional loss they strip them out and "underlying eps" gets top billing. Underlying earnings is what the company earns through its operations, so it's a good place to start for forecasting eps.

Watch out for "core earnings"; some companies like to emphasise this which all too often turns out to be "earnings minus some regular costs we'd like to pretend are one-offs in order to produce a higher headline eps".

[TheMotorcycleBoy]

So given trailing earnings, what do we do if, for example, we need the TTM earnings but we are missing last 5 months? In other words, it was that long ago when the firm last published results?

Which actually brings me quite nicely up to "Forward P/E or EPS" etc. What do analysts mean exactly when they say "Forward P/E"? Do they mean for the next quarter, half year, full year?

And of course, no time machines, no published results, what is an analyst meant to do?

Presumably, the analyst looks at a few prior, say 3, years and either deduces "the earnings are steadily rising by 5% per annum", or "over the past 3 years have risen first by 10%, then 8%, then 5%" and hence estimates a forecasted figure. (As SalvorHardin suggests, taking into account seasonal variations etc).

And lastly for "Forward P/E", does the analyst typically only forecast earnings ahead in time, and assume that the share will remain stationary? Or review any recent stock price trend and extrapolate out an estimate for this (which I have to say seems crazy to me).

M&M

[TheMotorcycleBoy]

Leading on from this, is another question. How much faith should be given to forward P/E figures? That is, compared to trailing ones which at least come from published reports.

[SalvorHardin]

Forward PE = Current share price / forecast annual eps

I have never seen PE calculated for anything other than one year's earnings. Bear in mind that when dealing with PE and eps "forward" and "forecast" mean the same thing (the next period).

For forecast eps analysts are attempting to predict eps by taking whateer information is available (e.g. past eps) and look at what's happening (e.g. sales, competitors, the economy), then coming up with an estimate. There isn't a special method; usually the analysts will forecast he entire profit and loss account based upon what they know. As you said, use previous increases, etc.

How much importance to attach to forward eps? Well spite the claims of some, financial and economic forecasting isn't scientific; far too many in the economics and finance professions suffer from "physics envy". It's educated guesswork, nothing more than that.

Estimates can be close (particularly with large companies which have frequent analyst briefings and loads of published information) or sometimes wildly inaccurate.

I rarely calculate forecast eps. Instead if I'm looking that closely I'll rely on something like the last year's eps and guesstimate by how much it will increase or fall. To me doing much more than that is likely to make my guess appear to be much more accurate than it really is. For large companies there will be a consensus analysts' forecast eps figure out there (just search for: company forecast eps).

As the saying goes, it's much better to be roughly right than precisely wrong.

[TheMotorcycleBoy]

Yes, I'm pretty much in agreement with you there, Salvor. I thought I had a recollection of some investment writer "preferring forecast figures where possible", but the cynic in me disproves of this type of guessimate - in fact it seems like rather a waste of time.

[TheMotorcycleBoy]

Since they usually give earnings per share, it seems a little daft to go back to first principles to establish that figure.

Actually they (Marshalls) do, look at the dark grey graphic, slightly to the right, and slightly above centre at their investor centre. It says P/E ratio 33.96, which is wildly incorrect, compared to my calculation of 17.05. By a factor of about 2......

Their figure is wrong, where as mine is correct.

M&M

[tjh290633]

Since they usually give earnings per share, it seems a little daft to go back to first principles to establish that figure.

Actually they (Marshalls) do, look at the dark grey graphic, slightly to the right, and slightly above centre at their investor centre. It says P/E ratio 33.96, which is wildly incorrect, compared to my calculation of 17.05. By a factor of about 2......

Their figure is wrong, where as mine is correct.

M&M

Found that. The half year figure says EPS are 13.24p for the half year. With a price of 440p, that gives you the number you first thought of, but only for the six months. You have to add the previous Half's EPS to get a 12 month figure.

TJH

[TheMotorcycleBoy]

Since they usually give earnings per share, it seems a little daft to go back to first principles to establish that figure.

Actually they (Marshalls) do, look at the dark grey graphic, slightly to the right, and slightly above centre at their investor centre. It says P/E ratio 33.96, which is wildly incorrect, compared to my calculation of 17.05. By a factor of about 2......

Their figure is wrong, where as mine is correct.

M&M

Found that. The half year figure says EPS are 13.24p for the half year. With a price of 440p, that gives you the number you first thought of, but only for the six months. You have to add the previous Half's EPS to get a 12 month figure.

TJH

Yes, I'd guessed that was the reason for the 2x discrepancy. Strange that their web page publishes such a value which actually makes their shares look less attractive at first glance, alongside the other points of interest: the shareprice and the marketcap.

I'm certainly more familiar with seeing PEs using annual EPSs.

But regardless of all this, my mine motivation for writing up this topic, was to debate, the pros and cons of using the TTM and the forecasted figures for PE and EPS. Since I was interested in considered forecasted figures it seemed prudent to at least consider how such figures are reached when there is not a ready supply of 12 months of immediately previous data for example here:

http://www.focusriteplc.com/investors/share-information

M&M