vendredi 30 mai 2014

GEA HY14 results

1) Intro

Note that the article is not self-supporting and will not make sense to a new reader (see my previous articles here 2012 , 2013).


GEA has published it's 2013 annual report and its 1stHY report:
- in the annual report the CEO warns clearly about a marked slowdown of sales in France, the need to focus on export markets, albeit at the cost of significant commercial investments and financial risks
- the CEO warns that it may impact profit margins
- 1stHY sales are down 25% compared to last year; down 50% in France, exports up 38%
- order book sharply down



I want to build a "worst case" scenario to decide what to do.
Besides, I previously compared GEA to Kapsch but Q-Free is maybe a better match (same size; Kapsch is much bigger, and builds but also operates whole/nationwide toll collection systems, a different business).
Q-Free 2013 revenue is the same as GEA.

2) What are GEA fixed costs ?

My current understanding of GEA business and financial situation is as follows:
- GEA designs the toll system (~100 engineers)
- I think it uses largely standard, off-the-shelf components (PLC, printers, card readers, audio and video systems, network components; fabrication of toll steel boxes is subcontracted, maybe they direcly manufacture some dedicated electronic cards) ; it's mainly integration and assembly, so no heavy equipment is needed (see below fixed assets). I wish I had the time to visit their plant near Grenoble.
 - ~100 people directly involved in production
- factory tests
- on-site installation
- maintenance and services

I think that if a downturn comes, GEA will be reluctant to significantly downsize its personnel (temporary staff aside) because:
- it would jeopardize future growth; GEA says in its annual report that the competence and stability of its people is important
- (supposition) not the style of this family-owned small company that thinks long term and not next year stock-options
- anyway long and costly (France labor market).

My main hypothesis is therefore that GEA has mainly variable costs, but that personnel expenses can be regarded as fixed costs, and it's a big part of expenses.

3) 2014 estimates

I've built the estimates with the preceding paragraph in mind and the following hypothesis:

- salaries: fixed cost
- all the rest: variable cost (not true of course)

- 1st hypothesis of 2014 sales~59 m€ (roughly double HY sales)
- 2nd hypothesis of 2014 sales~47 m€ (current order backlog)
- hypothesis that they will not needlessly build up their inventory and other costs
- I neglected amortization+depreciation in this simplified table, it's acceptable because it's generally low (see below)

Here is a table with data taken from the annual reports with my estimates.
Mat1ere = shorthand for "raw materials" expenses (in this case rather parts, components,...)
Autres achats = other expenses (go figure....)
Salaires = salaries
ROP = Operational result

I get a 2014 EBIT ~9-10 m€ in the (worst-case ?) scenario, a spectacular decrease compared to 2013.

But it would then trade at ~4 to 5x EBIT (for a quote of 87€), still a compelling valuation (because of the large net cash 60m€ for a 100 m€ market cap).


3) GEA competitive advantage
I've recently read this book The Little Book That Builds Wealth: The Knockout Formula for Finding Great Investments and liked it a lot; I will try to apply its concepts/ideas to the current situation.


GEA fixed assets are only ~ 1.2 m€ almost fully depreciated.
Much lower than Q-Free fixed assets though (see below). I have no explication for this.



It rents its buildings to another company...that belongs to the founding family/main shareholder, for what seems to me a reasonable price (see annual report).


Working capital is also very low; GEA pays its suppliers 30/60 days and negotiates payment terms with the customer (highway companies).

I've kept the actual denominations of the annual report (French version) for easy later reference.
Mat. 1ere = raw materials
en cours = inventory
creances clients = accounts receivable
dettes fournisseurs = accounts payable
dettes fiscales = fiscal debt
produits constates d'avance = prepaid income
BFR = working capital requirement



The capital employed (fixed assets + working capital) is thus very low.
About 5% of salaries can be considered as R&D costs, but are fully expended (R&D is not capitalized).
If I take assets at acquisition cost, add  5 years of R&D costs (arbritrary), ROCE is still very high.
Why is it not competed away ?

My understanding is that "local" actors (GEA in France, Q-Free in Norway, Kapsch in Austria,...) started early and small and ended dominating their national market (niche market).
I suspect that some kind of switching cost effect is at work there; for the highway company these actors provide a vital service, strongly integrated in the highway company process.
(Imagine the toll collection system down during summer vacations in southern France when half of Europe is en route from Netherlands/Germany/... to Italy/Spain/France...)



Besides from what I've read electronic toll collection investment pays for itself in less than 1 year.

So it makes probably little sense for a highway company to switch its ETC system from one to supplier to a cheaper competitor, unless there's a large price difference. Given the (small) size of the market, it's not just worth it for a competitor (again, a supposition). Additionally there is maybe (supposition) some notion of staff training (highway employees trained to do some basic maintenance on the toll machines) that reinforces this.

GEA benefited from this competitive advantage and an expanding market in France for structural reasons (highways companies going private investment cycle, see my previous article) but it is now apparently nearing saturation.

On the export markets, for new highways, I think that GEA has no clear competitive advantage (GEA "moat" is not scalable, lots of buzzwords here). Their strategy seems to accompany French construction companies (Bouygues, Vinci,...) on export contracts for new highways (see annual reports and also the red corner blog comments on GEA), but its unclear to me if it gives them some kind of commercial advantage on these markets.

5) Conclusion
I don't want to fool myself, and I'm the easiest person to fool.
I like GEA and I have invested some time (and  money) in this investment, but I want to remain cold and rational about this.
I'm certaintly not selling GEA at the current price.
I think a downturn is possible, but I think GEA is still an interesting investment with good long term prospects and favorable economics.
The stock is quite illiquid, and we may (or not) witness a spectular plunge if there's a knee-jerk reaction to weak HY results, maybe a buying opportunity.
I'd appreciate a reality check from outside investors.

vendredi 24 janvier 2014

GEA 2013 annual results

GEA published yesterday evening its annual results : link (in French only).
The presentation slides are interesting to get an idea of GEA business.
Previous posts on GEA: here and here.
The stock went up sharply today (+15%, 88.4 €).

A few comments
Sales are flat ; EBIT margin (operational margin to be precise) is up but GEA warns that it is exceptional and linked to the end of several contracts.


 The drop in backlog orders has -so far- not materialized itself. There's nothing in the press release or in the slides hinting at a future drop in sales ; the management insists on its export recent contracts. GEA does not seem to be much affected by the Ecotaxe Snafu.


The cash is boosted by a reduction of the working capital that was already apparent in the 2013HY accounts ; so again a one-off effect.

True to itself the management has decided to keep the cash to "stay independent" and "finance its investments and exports". However the dividend is up 40% (3.35 €/share).

The family owns 38 % of the capital. Michel Baule, an entrepreneur in polymers turned small caps investor, owns 15%. I see his presence (strong minority investor) as a positive development.

With 60 m€ net cash for a 106 m€ market cap, EV/2013 EBIT ~ 2 !
I see no excuse not to buy some more (and I did at the opening this morning).

A look at competitor Kapsch :


I did not spent enough time on this, but from what I understand the project-related part of the business can be quite volatile (problems in Poland and South Africa legal issues). So a good reminder that a mindless extrapolation of the past performance is dangerous...I guess it applies to GEA also.



mercredi 8 janvier 2014

2013 review : tops and flops

As last year, I'll review my past decisions and try to draw some lessons to improve my investing process.
So this is more a top of flop in terms of decisions, than in terms of performance.

The reason, as I explained last year, is that performance is driven by the process but also (at least in the short term) heavily influenced by luck and outside incontrollable factors.

I must also do a post on missed opportunities of which there were many. Let's Gowex (ALGOW) and Burelle (BUR) (both 3 baggers since I looked at them earlier this year, but without buying of course) come to my mind. It certainly isn't pleasant for my ego but I must see if I can improve my decision making here.


I'll review my previous posts more or less in chronological order

CIFE (INFE) : 58 € in Feb 2011, 50 € end 2011, 45 € end 2012, around 62 € end 2013.
INFE is still cheap because of its extra cash : 68 m€ market cap vs 65m€ NET cash (that stock screeners miss because it's not only in the cash and equivalent section).

However margins are currently paper thin.

The stock still trades at EV/EBIT ~1 even on these depressed margins.
Another issue is that liquidity has dried up.

So top or flop ? Maybe I got caught in a value trap ? I'll have to wait to end of the story to know for sure. But I'm certainly not selling at this low price.
Lesson learned ? maybe the time value of money aspect...undervalued situations can take a looooong time to play out -if ever-


SAM Outillage (SAMP) : 1st article ~32 € in Feb 2011, sold ~45-45 € end 2011. Buyout offer at 38+1.4 € special dividend end 2013 (at P/B~0.8, no intangibles). The float is now very low; a delisting is maybe in the works, probably at the same price.
So a good sell decision.
What made me sell end 2011 ?
I sold at about EV/2010 EBIT ~ 7  but 2010 PER of ~16 and based the following considerations (according to my notes and what I can remember) :
- I was well aware of SAM business characteristics (low margins, cyclical, no concurrential advantage vs other tool makers
- SAMP had a large cash position but had decided to make some major acquisitions, with the associated uncertainties
- and the stock was nearing its highs ("technical analysis", shame on me).

Lesson learned n°1: maybe that "low quality" (not meaning here SAM management, but simply the sector of activity/concurrential position) companies must be bought super cheap and sold cheap. And also maybe not to be blinded by one valuation metric (EV/EBIT in my case): see lesson 2

Lesson learned n°2: SAM net cash has gone from +5.2 m€ end 2010 to -0.4 m€ end 2012 due to new outside companies acquisitions. Sales are up 30% but the net margin has been (for now) divided by 2. A good reminder that extra cash does NOT belong to minority shareholders and I should be careful with that.

Total (FP): 1st article ~32 € August 2011, sold some around 39 € end 2011. Around 45 € end 2013, not counting the hefty dividend. Why did I sell ?
My investment thesis was that I was buying Total for its dividend. My sell decision was inconsistent with that.

Vivendi (VIV): 1st article ~16 € sept 2011, mostly sold around 18 € -19 € recently (see post here, based on sum of parts analysis).
Some local perspective for foreign readers.
As I feared in my previous post, price war has restarted around 4G offers end of this year (which were supposed to improve the revenue per user of operators). Free mobile said they would offer the 4G for...free with their current subscriptions. Only issue is that their network coverage is very very limited. But the damage (in the customer mind) is done.
Bouygues responded by cutting its 4G subscription prices and I expect Vivendi (SFR) to follow. Inflammatory exchanges in the press. Bouygues CEO threatened to cut Free (Illiad) primary source of income by starting a war on fixed internet prices. Free stock took a 6 % plunge this same day.
Vivendi has sold major divisions, and announced that SFR would be spin (spun ?) off  (but at what price ?).
So far I'm still satisfied about my decision. I've kept a small residual line to keep me interested in the company and see how it turns out.

Tessi (TES) : 1st article~62 €, around ~90 € now. Will do a separate article later.

Groupe Crit (CEN) : 1st article ~20 €, averaging down ~14€ and 11€, sold some around 16€. Stock around 31 € now.
So a big nice clear flop.

Lesson learned : taxes !
The apparent tax rate (by that I mean the tax rate calculated frm the income statement and reported in Crit annual reports) has varied widely with time.

The 2009-2010 jump I was aware of ; it was something to do with the calculation of the CVAE tax, moving from being deducted at the level of the operational result to the net result (or so says the annual report). Crit was trading at a very low EV/EBIT ratio but I was aware of and correcting for this tax effect.

But then (after I sold) the 1HY2013 reports an apparent tax rate of 33% vs 65 % for 2012 !  I suppose that this is linked to the newly installed taxe rebate for low salaries (CICE). I've not seen an effect so marked for other companies I follow, I think this is related to the nature of Crit business (interim services, lots of (low) salaries, low fixed costs).
Besides the company has apparently done a good job of diversification in the US and the interim services sector has improved.
So clearly my mistake can be classified in the "outside circle of competence" folder ; it's worthwhile to spent more time on the taxes section.

LaCie (LAC): 1st article ~3.5 €, 1st buyout offer from Seagate at 4.05 €, 2nd offer at 4.50 € (I did not sell then either). Seagate has reached the 95 % treshold and the stock has been delisted (compulsory sale for minority shareholders like me) end 2013.
Lesson learned: maybe that a company can be taken private at a low price if there is no strong minority shareholder (mutual fund,...) to protect you.

FFP (FFP): 1st article ~25 €, around 43 € now. At that time my investment thesis was that it was a bet on Peugeot not going belly up with a downside protection guaranteed by the value of the other participations of FFP. Peugeot itself is up 74%, but it's been a wild ride.
I'm quite satisfied with myself on this one. I sold some around 42€, just to trim my exposure. The question of whether there is still upside potential warrants another study and a separate post.
Lesson learned: should do more investing based on balance sheet/sum of parts analysis.

GEA (GEA): 1st article ~65€, around 76 € end 2013. Still trades at EV/EBIT ~ 2. I continue to be puzzled by the stock undervaluation, even if sales and order backlog are quite volatile. Apparently the minimal financial communication is not very appreciated either (Small caps confidentiel link). Michel Baulé, a noteworthy entrepreneur turned small/mid cap investor, now owns 15 % of GEA capital.

Among other buy decisions that have turned out well are IGE+XAO (IGE) or Gerard Perrier Industrie (PERR), both high quality small caps, that I've net yet covered in the blog (but the value and opportunity blog did).

The post is getting quite long and possibly boring so I'll stop there.

And finally, best wishes to everybody.

jeudi 19 décembre 2013

2013 review

The year is nearing its end, so it's time to reflect on 2013.

My porfolio is up approx 35 % (defined in my case as the internal rate of return, XIRR function of excel), so I can't hardly complain.
However some of the Small caps value funds I follow are doing much better: see this link (video in French):
Independance et Expansion is up ~60 %, HMG Decouvertes up ~40 %, Sextant PEA ~37 %, Moneta Micro entreprises (closed to new subscriptions) ~ 36%.

As a consequence, I've had a hard time recently finding interesting stock picks (and the blog has remained quite silent). Same thing with convertible bonds, risk aversion is sharply down whereas it was possible to find a good yield on some thinly traded convertibles only 6 months ago.

This may be compounded (in France) by the creation of the PEA-PME, a tax shelter (a rarity these days) reserved for European small cap stocks.

It also seems to me (but I haven't bothered to quantify this) that IPO's are back in Euronext Paris, especially for med-techs. The small caps are also quick to react to news (publications,...) whereas a few months ago the market was quite inefficient and it was possible to buy a few shares before the news had spread.

My favorite gurus have issued what I consider to be warning signals (this applies to the US markets, more richly valuated than Europe, but still...)
Hussman funds
Howard Marks
Jeremy Grantham


In a next post, I'll review my tops and flops and try to draw some lessons for next year.
And Happy Christmas to everybody.

mercredi 2 octobre 2013

Nexeya buyout offer

I own some shares of  Nexeya (ALNEX), bought between 5 and 8 € in 2012.
I never made a post on Nexeya but Nate at oddballstocks did.
Nexeya is currently making a tender offer at 12 € for minority shareholders.

Nexeya released yesterday the details on its offer (link, in French only I fear), at the same time as its 2013 financial accounts.

It's interesting to look at the offer for 2 reasons :
- the actual decision to sell or not,
- it's very instructive to look at the independant expert valuation and methodology, to improve mine but also to have an idea about what to expect in future comparable buyouts.

The document is 76 pages long, I'll try to make a shorter post !

The independent expert is a member of Crowe Horwath network.

The independant expert uses DCF and comparables analysis.

Any approach based on book value is judged irrelevant by the expert for a company like Nexeya. The offer values Nexeya on a P/B ~0.7 (based on 2013 HY accounts) or P/B~0.9 (based on 2013 final accounts, after a ~20 goodwill depreciation has conveniently depressed book value).
1st lesson: never assume that the minimum buyout offer will be at book value (even for a company that's been profitable for the last 7 years, median ROE~10 %).


DCF analysis
I show only the final table:


The main hypothesis built into this model are :
Cost of capital 12.2 %  (based on WACC stuff :link); I routinely use 12 % as my "base rate". So no bad surprises here.
Projected EBIT margin 7.3 %, which is more that the average past margins and 2010 peak margin at 7.1 %. So again a slightly optimistic but reasonnable assumtion.
Modest future growth, again a reasonable assumption.


Then comes net debt :

3.2 m€ for apparently unfunded pension liabilities (after taxes), a good reminder to not forget those in my own evaluation.

However, if we have a look at the most recent published accounts (annual accounts, June 2013), and if I'm not mistaken
short term debt 5.2 m€, LT debt 7.6 m€, cash 16.0 m€ => excess cash 3.2 m€ vs
1.1 net debt according to the independant expert !!!!!!

Where did all this cash go in just 3 months ?
I have not found the answer. I suspect that it is used to finance the buyout, thus depressing the tender offer price. Neat, isn't it ?
Correcting for this only, without changing the rest, would make the DCF estimate go from 12.5 €/share to around 13.5 €/share.

Comparable companies


This graph plots EV/Sales vs EBIT margin, using 2013 expected results.
You can always discuss if the companies are actually comparable (Steria provides information technology services, not quite close to Nexeya), but that's not the point.
Based on this "regression analysis", and Nexeya 7.3 % EBIT margin, the expert computes a "fair"value for Nexeya.
Then, because Nexeya is a smaller company than the mean of the sample used here, the expert applies a 15 % discount on this :


No fundamental reason other than an empirical observation on market cap size effect on valuation.
Let's have a look at the comparables market cap


This one reminds me of the "Bill Gates walks into a bar" example about the distinction between
the mean and the median value of sample. With Thales in the sample, we can be quite sure that Nexeya market cap is smaller than the mean.

Another issue (but I don't feel like going on, it is late) are the projected results/margins used for the comparables ; I checked some and I found significant discrepancies between projected and actual known 2012 data, often leading to a higher projected price for Nexeya "fair" price.

Conclusion

I've not made up my mind yet. I need to cool down a little.
This confirms my first impression about a lowball offer.
However I'm not sure I want to stay invested in a company where the managers interests are evidently not aligned with mine, and maybe severe liquidity issues in the future. Besides, I'm not sure if there is some dilution around the corner or not.

I want to make some further observations.
Given the path of Nexeya stock price (IPO at around 17 €, buyout at 12 € ~seven years later), this reinforces my rule of thumb to never participate to an IPO, which is a pity because I view it as one of the few useful functions of the stock market for the "real" economy.

This also justifies why you need a large margin of safety compared to a "fair price", maybe even more so for small caps maybe given this kind of tender offers, especially when no strong minority investor (independent mutual fund, activist shareholder,...) is present at the capital.



lundi 16 septembre 2013

Stockopedia review


Version Française : à venir.
 
Intro

I’ve been recently offered a trial subcription to Stockopedia and it gave me the opportunity to test their stock screener.

This is my review. I’m of course grateful for this gift, however I’ll try to be unbiased. Being French and the Stockopedia founders British, I’ll also try not to add to our national reputation of arrogance (largely unfounded of course).

The investment process looks more or less like this (value and opportunity blog copy/paste)
A) Stock screening & quick analysis
B) Deeper Analysis
C) Buy decision (or not buy)
D) Sell at some point in time
Obviously we’re talking about step A).


Currently I’m using a personal “database”, basically just a bunch of Excel sheets. Every time I look at a stock, I make a spreadsheet, and I try to update it (with new financial releases) when I look again at the stock a few months/years later.
Why Excel ? I already have the software, I know how to use it, you can draw graphs, make some personal calculations, keep some notes on the company you’re looking at (for instance why you invested/did not invest in this company ; it’s always funny and self-sobering to look back at your decisions and thought process a few months/years later).
Currently this “database” is around 12 Gb (and clearly Excel was not meant for that, I do have some crashes from time to time) and the manual updating process is quite (meant as a British understatement) laborious. And the other hand, it forces me to read the annual report, look at the accounts, try to make sense of them, in short it forces me to make the transition from a stock to a real-life company. 

Crash test of the stockopedia stock screener


I do use stock screeners from time to time to generate ideas, because my database is not exhaustive and I can easily miss opportunities.
There is much much less coverage and screeners for the French market compared to the US and UK markets. The only free one I know is the FT screener (link), and it’s pretty rudimentary, but, hey, it’s free.
The stockopedia screener is neatly presented and is pretty intuitive to use (and anyway who bothers reading manuals, these days?).


A wide choice of ratios and financial metrics (ratios based on earnings, sales, FCF, debt, growth,…) is available and furthermore you can combine them. Actually you’re offered with so many choices that I felt kind of overwhelmed at first. But I think the designers have foreseen this and some simple “push button” screens following a well-known strategy are also built-in. The nice thing is you can start from this base and tweak it to build a custom screen.



Once you’ve selected your database, you run the stock screener and voilà !

A feature I like is that you download your results in excel or other file (pdf, csv, xls,…) for further analysis offline.
So let’s do a crash test on the French market; to start I used a simple EV/EBIT ranking (actually the inverse ratio = earnings yield, defined by stockopedia as operating profit / enterprise value).
Here is what you get (only part of the list).
 



On this list, there are many companies I’ve invested in or looked closely as buying opportunities, or covered by other value blogs: ADL Partner (see article here), GEA (see article here), STAL (see article here), Maisons France Confort (MFC, see article here, or ennismore funds description here), just to name a few.
I checked the numbers against mine and published results and found them generally accurate, so it means that the database is reliable (more on that below). The database seems also to cover the French stock exchange pretty exhaustively.

Minor bugs


It’s interesting to look at the top companies of the EV/EBIT list.

Promeo sold a division in 2012, thus boosting the operational result. The actual Promeo financial statements make the distinction between current operational results and operational results but the screener uses only the operational result, hence the artificially low EV/EBIT ratio.


Same thing for Richel Serres (company makes greenhouses). It sold a major subsidiary last year, thus boosting the net result, but not the operational result, or Interparfums (perfume maker ; current operational result 58  m€ vs 213 m€ for Stockopedia ; the difference is due to an exceptionnal settlement after the termination of a contract/licence with Burberry).


I was NOT surprised NOT to see CIFE (ticker INFE) high in the list. The company has roughly double the cash reported by stockopedia, but it’s hidden in the non-current assets (see my old article here), and a stock screener can’t know that. 

However I don't want to make the impression that the database is flawed. For the other companies I checked, the numbers were correct and up to date.

Other features

Of course you can build a much more elaborate screen, blending price/value indicators (earnings yield,…) vs “quality” indicators (whatever that means for you : high ROE, ROA, ROCE or operating margin or net margin, or FCF yield, just to name a few).

Stockopedia has a built-in pretty elaborate ranking engine. As I said previously you can always use this as a template and build your own. Here’s a screen shot :


I own a lot of shares of companies in this list  (Infotel, Gerard Perrier, Linedata, Neurones, MFC, ) and have seriously looked at most of them as possible buy candidates (LSS, PUS, LCO,...), so from my point of view (meaning my own buying criteria) , I find the results rather impressive.

I'm also surprised by the large number of companies on this list in the Computing Services Sector (INF, LCO, OSA, SOP, NRO) ; I don't know if it's the result of an actual undervaluation or the result of the choice of screening criteria.
I know that there's a sizeable difference between operational and net result for these companies because of taxation issues (specific to these companies that pay lots of salaries but with little supplies/raw material costs), so some careful double checking is needed here.


Conclusion

To sum it up I think the screener is a very useful tool that gives you a solid list to narrow the choices and start digging deeper.

I've also been impressed by the reactivity of the Stockopedia team to answer requests and fix bugs.

I wouldn’t rely on the screener results alone to make the decision to invest, and I don’t think it is Stockopedia goal either but it’s a tremendously time-saving tool.

I think also this is very useful when trying to expand your investment horizon outside your home market. I remember a Buffet article / interview saying that at the beginning of his career he started looking at all the US stocks one by one alphabetically. I doubt I would have the patience and available time to apply this method for the UK or German stock markets, which are pretty large.

dimanche 30 juin 2013

GEA : the signal or the noise ?

The title refers, of course, to the book by Nate Silver, that I'm currently reading.
The Signal and the Noise

I like the book a lot (even though the chapters about baseball are hard to follow if you don't know the game) and recommend it.

GEA has recently offered me a good opportunity to think about how to filter the signal from the noise.

The facts (as objectively as possible)
I covered the fundamentals of GEA in a previous post : here and here.


Recently GEA published its HY results (here, in French only).
Earnings are in line with the previous year.
Cash now amounts to 54 m€, for a market cap of 88 m€ (73 € last quote). 0 debt. So EV/2012 EBIT ~2, super cheap for a company with apparently strong margins in a niche market.


So I was tempted to add to my position (and I did, much higher than 73 € though).

But :
At the end of the news release, GEA announces a 57 m€ backlog, a dramatic decrease from 75 m€ last year, even though GEA says at the end of the release that new orders have been booked since, notably in export markets.

The stock started a rapid descent the next day :



I treated this as noise but the "market" apparently thinks otherwise, and it's causing me sufficient concern to revisit my decision and the process (or lack thereof) behind it.
I'll write it down here and it will be interesting to revisit here in a few months.

I guess that most of my readers will have been in a more or less situation and would like to hear their thoughts on how they treat such situations.

So, signal or noise ?

On this graph I've plotted the backlog and actual sales of GEA for the last few years.
It makes sense that backlog and sales are highly correlated.
Last time the backlog dropped around 2007, the sales did follow a few months later, although they dropped much less. And a one occurrence sample makes for poor statistics...


Hypothesis 1
Recent order book drop is a fluke; current valuation is a steal. Sellers are wrong. Buy.

Hypothesis 2
Recent order book drop is highly significant. It's an indication of a large future drop in sales. I'm the idiot for trying to "buy the dip".
After all, many recent infrastructure projects have been recently cancelled in France, not good for GEA business.

Moreover, management does not say anything about this, which shows a serious lack of communication and deserves a higher valuation discount rate.

If we correlate the order book and the sales, we could expect future sales to be around 50 m€.
It also seems possible to assume that the margin will decrease and revert to its long term mean (around 11%).

This means an EBIT of around 6 m€, so an EV/EBIT ratio of around 6, not exactly a nosebleed valuation either, and certainly not a reason to dump the shares as brutally as happened ?
At the last shareholders meeting, some shareholders tried to have a special dividend voted, but the owning family blocked this. So maybe some are considering that the large excess cash of GEA is locked out, and that it deserves an additional discount ?

Conclusion

I can't really make sense of this brutal drop, but I'm maybe missing something or fooling myself ? And after all value investing is not about trying to make predictions about next year earnings, but rather putting the emphasis on valuation. I think GEA is a bargain at current prices, but time will tell.