Stock analysis and other stuff: July 2014

Monday 21 July 2014

Director purchases and sometimes lazy longs ...

Monday 21st July 2014

A common misconception used by longs against short sellers is that a short argument can be negated by the fact that a shorted company's management may have bought stock (typically subsequent to a short attack). This is not always so and can be a dangerous (and lazy) assumption. One of many instances where it proved a mistake is in the case of Connaught.

In March 2010, I was four years in to my six year stint in sell side equity research. I'd issued a bearish Sell note on Connaught Plc (CNT). At that time CNT was capitalised at £400 million, having fallen by c. 35% over the prior few months from its all time high market cap of c. £615 million. During the decade leading up to this CNT's share price had risen from 30p in 1999 to north of 400p per share in 2009. CNT was a decent sized, well covered (not to be confused with well researched) company. The group provided repair and maintenance services to the UK's Social Housing market, as well as compliance services, such as health and safety.

The analysis in my sell note principally focused on CNT's woeful record of cash generation in relation to its reported profits, its increasing capitalisation of software related costs, the frequency of 'one-off' restructuring charges, and its rising use of debt for acquisitions. I reckoned CNT was overstating profits by at least 20%, so I set my financial forecasts at around that level below market consensus. In the event, this was far too optimistic. CNT went into administration six months later in September 2010, seemingly on the back of some form of fraud.

Several things surprised me during those six months between March and September 2010.

Firstly I was amazed at the vitriolic response by CNT's management to what I thought was a fairly straight forward note. All I had really highlighted were facts that had led CNT's business and financials to where it was at the time. CNT's CEO, Mark Tincknell, issued a particularly pernicious email to his employees, institutional investors and some press concerning my research note. Further, CNT was so popular amongst institutions, and the sell side, I think there was a general perception that I'd gone mad to have issued a bearish note.

I was also surprised by the market reaction.

CNT's share price fell by 15% in the days following publication of my note. But then the shares staged a strong c. 33% rally in the months to end of June 2010. The rally was seemingly prompted by the arrival of a new Chairman, Sir Roy Gardner, a number of brokers coming to CNT's defence issuing 'strong buy' recommendations, an upbeat set of interims in April 2010 where the CEO commented "there are no big surprises out there", and director purchases.

Director purchases.

In the weeks following April 2010's interims, CNT's directors purchased stock. Robert Alcock bought 10,000 shares at 304p, the CEO, Mark Tincknell purchased 335,399 shares at 309p (on the face of it an impressive £1 million worth), and Sir Roy Gardner picked up 158,808 shares at 315p. Poor Sir Roy. No really, poor Sir Roy. Half a million down the plug hole. Within six weeks of their purchases CNT issued a whopper of a profit warning and the shares fell by 67%. Poorer Sir Roy stepped in again. He bought a further 50,000 shares at 112p.

As fore-mentioned, within another few months or so of the whopping profit warning in late June 2010, CNT was a bust seemingly on the basis of some form of fraud. There is no doubt in my mind that the rally in April and early May was partly attributable to the perception of director purchases.

Given Sir Roy's late arrival to the fiasco, he was unlikely to have been involved in any improprieties. All he seems to have been is a pretty poor judge of character and financials. But why would the CEO have sent a further £1 million down the Swanee? One answer may be that in the year prior, Mark Tincknell (who at that time was the Chairman), dumped 650,000 shares on the market at 355p per share. So in context he bought less than half his holding back when things began to go sour. I also recall that in October 2009, almost at the peak of the share price, the former CEO, Mark Davies, sold close to £7 million in stock, while the former FD, Stephen Hill, sold £3.8 million worth of shares. Talk about timing!

Since then I've been wary of director purchases in situations where a cloud hangs over the underlying business, especially when purchases are in repeated flurries and small beer compared to prior director sales.

Connaught Plc's share price
Source: Bloomberg

Connaught Plc director purchase and sale of stock
Green = purchase
Red = sale
Source: Bloomberg

Disclaimer: The information, discussions or topics referred to on this blog should in no way be considered “advice” to buy or sell anything. The information which may be referred to is freely available in the public domain and where required the source of information is referenced to for verification. While every effort has been made to ensure the veracity of any information contained within this blog, the author accepts no responsibility for the accuracy of any information contained within this blog or for the sources of information which may be referred to. Readers are responsible for their own actions and interpretation of the information contained within this blog.

Avanti (AVN) ... gravity and reality

Monday 21st July 2014

In light of Avanti's (AVN's) share price move since its trading update from Friday last, there is an element of the horse having bolted. Nonetheless, as I still reckon any would be buyer mad to pay a high two digit penny sum for an AVN share, the following may explain my position. I am short Avanti.

Firstly, there is a stark contrast between last year's statement and this year's. Where are the KPIs and the business update that we had last year? Going by previous RNS releases, it's very unlike AVN to miss the opportunity to talk up the business.
"Revenues are expected to be in line with market expectations in the range of $64 - $65m". As the chart below highlights, revenue expectations received their latest cut in May last, and have been on a downward trajectory since 2014 forecasts were anticipated to be over $300 million in late 2011.

Avanti 2014 consensus revenue expectations
In line with the cut in May 2014, but on a downward trajectory since 2011
Source: Bloomberg

What the revenue expectation means is that AVN achieved quarterly revenue in Q4 of c. €25 million, which represents a 79% increase in revenue from Q3. This seems like a great result. However, it would be illustrative to learn how much of this revenue jump relates to the $15 million Government contract announced at the time of the interim results?

"In Government, the launch of Avanti Government Services has made an impact on our ability to promote ourselves as the high security provider of choice. We were awarded a $15m project to build a high value-added network in one African country and are confident that more strong government business will be forthcoming shortly."
The Government contract is certainly revenue but it's not ongoing revenue from the sale of satellite capacity that is repeatable. It's probably also not as profitable as the sale of bandwidth and would certainly help explain why profitability continues to be so slow.
AVN announced that PBT would be lower than consensus with approximately half the variance resulting from bond refinancing costs of $7 million. This suggests that profits will be lower than expected by $14 million. As AVN's joint broker, Jeffries, rightly highlighted in its note from Friday last, the refinancing costs would have been known by management for some time and most likely at the time of the group's $150 million bond placing in early June 2014.
Allthough AVN suggests that its PBT will be lower than expected by $14 million (bond refinancing being half of this at $7 million), Jeffries has cut its 2014 PBT forecast by a whopping $58.5 million, from a previous $22.8 million loss to a $81.3 million loss.

Avanti 2014 consensus PBT expectations
AVN suggested PBT would be lower by c. $14 million but house broker Jeffries cut forecast by $51.5 million
Source: Bloomberg, Jeffries

Cash is reported as being a very healthy $195 million. But AVN raised $157.5 million through a bond placing in Q4. So without that, it would appear that AVN would have been close to running out of cash. It also helps to illustrate why very little of the jump in revenue has failed to turn into cash or profit. Without the bond raise, cash would have been:

$195 million (year end cash figure)
less $157.5 million (amount received on bond issuance)
= $37.5 million (cash at 30 June without the bond issuance)

Hence, cash consumed during the quarter would be:

$65 million (cash balance at end of Q3)
less $37.5 million (what cash balance would be without bond issuance)
= $27.5 million

$18.5 million of this went on bond interest paid on 1 April
So
$9 million of cash was consumed during Q4.
Some of the $9 million in cash consumed would have been HYLAS 3 capex related. Brokers forecast $25 million per annum, so say $6.3 million per quarter. This still implies that there was a $2.7 million cash outflow after capex and despite the bumper rise in revenues. As usual, when the accounts come out it will be interesting to observe the quality of those revenues.
Without the fund raise and with the above rate of cash burn, AVN would likely have been down to its last $6 million on 1st October after paying the bond interest (6 months on $370 million and 3 months on $150 million). So the raise looks very much like a rescue raise.

This all begs further questions ...

How could AVN have told the market that it was trading in line with expectations in the Bond press release announcement?
What will Moody's do now that an anticipated 10x debt multiple when the bond was issued is now an infinite multiple of that loss?
Will AVN proceed to buy the 4th satellite or just use the extra $150 million to get it through (if that day arrives) to becoming cash flow positive?
What is the Chairman, Paul Walsh, doing these days? How involved is he in all the obfuscation?
Will the high yield bond market be open to AVN in future years, because it appears that there's a lot of new financing and re-financing to come???

And another thing ...
No wonder Caledonia (AVN's 2nd largest shareholder) started selling recently and before the update. The question is, having sold c. 2.2 million shares and having around 13 million left, are they done selling?

Avanti share price
Source: Bloomberg

Friday 18 July 2014

The Pig and Pork Scam ...

Friday 18th July 2014

'Pig and pork' is the process whereby Pig Limited buys from Pork Limited, apparently at arm's length when in fact there is undisclosed or unclear common ownership of both companies. The result is that profits emerge in Pork Limited so that Pork Limited's financiers are conned into doling out more cash/finance, in effect to both companies. The auditors cannot systematically pick it up; especially if Pig Limited and Pork Limited appoint separate audit firms. And the only financial cure for both companies is that Pig Limited ultimately disposes of that which it has purchased at a profit after covering the warehousing costs incurred. Needless to add, this cure is not usually to hand in time. The practice is fraudulent.

The definition above was provided to me by the Varlet, so I thank him for it.

Now, how might this be applied in practice? Here is a fictitious example ...

Let us suppose there is a company listed on London's AIM market, called Pork Limited. Pork Ltd has operations across many countries but is head-quartered in that bastion of financial rectitude that is Sofia, Bulgaria.

Historically, Pork Ltd has reported growing revenue and profits but generated little in the way of cash. The cash was all tied up in debtors. However, debtors were not settling their accounts because the entries related to fake revenue. Fake revenue and profits continued to be reported. Debtors continued to pile up. Cash was still not generated and so Pork Ltd continually tapped the markets for finance.

As a few years passed, the lack of cash generation and piling up of debtors became embarrassing. Debtors were being shoved in such places as trade receivables, post dated cheques, and amounts recoverable on long term contracts. So a solution had to be found. Amongst no less than say 13 subsidiaries, Pork Ltd had a wholly owned subsidiary called Pig Ltd; also head-quartered in Sofia. Pig Ltd was to be divested and the fake revenue and fake debtors were to go with Pig Ltd. However, Pork Ltd could not find a buyer for Pig Ltd (fake businesses are difficult to sell), so it sold 51% of Pig Ltd to Pig Ltd's management and financed the transaction for them. Strangely, Pork Ltd handed over, let's say €5 million in cash to Pig Ltd's management, and the terms of purchase would be that Pig Ltd's management paid, say, €10 million in instalments over say four years back to Pork Ltd for their purchase.

But Pork Ltd was still largely a fake business and with still largely fake revenues and fake profits tied up in high fake debtors. And with the promise to the market that fake revenue would go higher, real cash was urgently in need. So it tapped the market for its biggest equity raise ever, say for €20 million, with the use of proceeds supposedly for acquisitions, and working capital to support its impressive but fake growth. A little later in the year it took on a further, say €18 million, by way of debt. Nonetheless, Pork Ltd was still committed to demonstrating revenue and profit growth and needed to show cash coming in.

Meanwhile, Pig Ltd took the opportunity to also raise debt. It raised say €16 million. So with the €5 million received from Pork Ltd, Pig Ltd had around €21 million in cash available. Free from the burden of having to demonstrate fake revenue to the AIM market, Pig Ltd's revenues dropped by say 40% in the year following divestment. This was all largely in one division, Pig Ltd's ~~software~~ butchery products business, where revenue fell by say €15 million or say 75%. Oh yes. €15 million.

Back at Pork Ltd, it couldn't find any buyers for its butchery products. But it had led the market to believe that it had enough buyers to purchase say €20 million of its butchery products. So Pork Ltd got into contact with its now associate, Pig Ltd.

Pork Ltd highlights to Pig Ltd that as it has cash, why doesn't it buy butchery products from Pork Ltd? After all, Pig Ltd can always capitalise the cost of its purchase of Pork Ltd's useless butchery products on its balance sheet, and still report good fake profits. Pig Ltd's management keen to help out and mindful that they still owe Pork Ltd maybe €8 million, agrees. It might appear strange that Pig Ltd's butchery related revenues have collapsed and that it still spends and capitalises an even greater amount on butchery product development cost, say €12-18 million. Oh yes. €12-18 million. But it is in Sofia, and few will ever look at its accounts.

To much fanfare, Pork Ltd's butchery product revenues grow by €20 million. Joyfully it collects some cash. Unbeknownst to the market most of the butchery products were purchased by Pig Ltd and the cash received from Pig Ltd. The Pig and Pork Scam continues ...

Thursday 17 July 2014

Volatility looming? ...

Thursday 17th July 2014

I bought the VIX earlier. After all, it's been a while since a decent spike. The mid to high teens looms large ...

Volatility Index
Source: Bloomberg

It's been longer still since a surge into the 20s ...

Volatility Index
Source: Bloomberg

Wednesday 16 July 2014

Globo (GBO) ... a peek at the Greek

Wednesday 16th July 2014

Globo (GBO) divested 51% of its Greek operation, Globo Technologies (GT), in Dec 2012.
GBO values GT at c. €23.7 million.
57% of GT’s net book value appears to be related to capitalised intangibles/software.
17% of GT’s net book value appears to be trade receivables over one year past due.
GT’s revenue declined by 29% in 2013.
GT's software related revenue fell by €14.480 million or by 75% in 2013.
GBO's Enterprise and Mobile software project revenues rose by €17.781 million in 2013.
GT reported a €12.529 million cash outflow in 2013, largely on intangible/software items.
GT’s net debt has risen to €13.731 million in 2013 (2012: €0.780 million).
GT trade receivables and revenue recognised under IAS 11 is 101% of 2013 revenue.
36% of GT’s trade receivables are over one year past due.
GT has not reported positive operating cash flow for two consecutive years.
GT's 2013 free cash flow was negative c. €12.867 million.
GT is held on GBO’s balance sheet with a value of €11.625 million (2012: €10.464 million).
GT's management still owe GBO €9.7 million.
I am short GBO.

Globo’s (GBO) 49% stake in Globo Technologies S.A. (GT) is held on Globo’s balance sheet with a value of €11.625 million as “Investments in an Associate”, while €9.700 million is due from GT's management as “Proceeds from disposal of a subsidiary”. A total of €21.325 million is attributable to GT, which is just over 15% of GBO’s net asset value for 2013. Other than an update from September 2013, little has been heard of GT since GBO’s Greek operation was divested on 3 December 2012.

Globo 2013 accounts: Investment in Associate

Globo 2013 accounts: Proceeds due from Globo Technologies

GBO sold 51% of its subsidiary, Globo Technologies (GT), to a company called, Zipersi Consulting, owned by GT’s management team. GT was sold for €11.2 million. This was deferred consideration. In fact the payment schedule seemed particularly favourable, with only €2.0 million of the deferred consideration due over the first two years and €9.2 million due from December 2014, through to a final instalment of €3.7 million due in December 2016.

In addition to the preferential financing terms, according to GBO's 2012 annual report, in note 15, while an initial €400,000 was received by GBO as a first receipt of consideration due from GT's management team, it would also appear that €7,061,000 in cash and cash equivalents went to GT. This resulted in a net cash outflow on disposal of GT, of €6,661,000. This is an altogether odd flow of cash for a disposal. Usually one would imagine that the cash flows from the acquirer to the vendor and not the other way around, albeit with an understanding that it flows back over the years to come.

Net cash outflow upon disposal from Globo to Globo Technologies
Source: Globo 2012 annual report

Flow of cash from Globo to Globo Technologies
and anticipated deferred consideration schedule from GT's management team
Source: Globo 2012 annual report

One may ask why GBO elected to sell 51% of its stake in GT? If the business was bad, then why not sell all of it? If it was a good business, then why sell it at all, or even as much as 51%? Especially on preferential financing terms to the management team buyers?

The upshot of all this was partly aesthetic in that a boat load of receivables prior recorded on GBO’s balance sheet, were replaced by a relatively anodyne “Investment in associate” and “Proceeds due”. Indeed this was alluded to in the accompanying divestment announcement:

“In addition, the exclusion of assets and liabilities (including debt) from the Group’s balance sheet will provide significant additional visibility to the investor community on the Group’s international operations and financial performance.”

Globo updated the market on 23 September 2013, as to its own and GT’s performance. At that time, GBO’s CFO, Mr Dimitris Gryparis stated:

“The divested company Globo Technologies S.A. is out-performing expectations and trading ahead of forecasts. For the six months ended 30 June 2013, its revenue increased by 233% to €13.64 million (H1 2012: €4.09 million). It has a pipeline of public and private sector contracts of more than €10 million to the end of the year.

Profit after tax reached €1.49 million, with that attributable to the Globo Group, as a 49% shareholder, being €0.7 million.”

GT’s 2013 full year accounts can be found here. They highlight that revenue in the full 2013 year was €25.224 million, down from €35.701 million in 2012. There are several things to note here:

This is a dramatic year on year decline in revenue. Revenue fell by 29%.
If, as GBO’s CFO highlights, GT achieved €13.64 million of revenue in H1 2013, as against €4.09 million for H1 2012, then this means that GT’s H2 2012 revenue would have been €31.611 million. This would imply that while GT’s H1 2013 revenue may have been 233% higher as compared to H1 2012, that H1 2013 revenue was 57% lower than revenue achieved in H2 2012. Further, at €11.584 million, H2 2013 revenue would be 63% lower than H2 2012 revenue. See GT's revenue trends as implied by Globo statements and GT's 2013 accounts in the chart below.
GT's revenue decline was attributable in the main to a €14.480 million fall in software applications revenue. This fell from €19.209 million in 2012 to €4.729 million in 2013. A drop of 75%. Incidentally, in the same year, GBO's Enterprise mobility licences & subscriptions and Mobile software projects business lines were going great guns. They grew revenue a combined €17.781 million in 2013. As highlighted further down, while GT's software related revenue plummeted, its cash outflows on capitalised software related items went considerably higher. Why was this when revenues were falling so sharply?

Revenue of Globo Technologies as implied by Globo statements and Globo Technologies accounts
Source: Globo, Globo Technologies annual reports

The quality of GT’s revenues is also worth considering. GT’s accounts highlight that revenue rose and fell as follows in the chart below during 2010 to 2013:

Globo Technologies revenue: 2010 to 2013
Source: Globo Technologies annual reports

GT’s trade receivables were as follows during 2010 to 2013, highlighted in the chart below:

Globo Technologies trade receivables: 2010 to 2013
Source: Globo Technologies annual reports

GT’s revenue recognised under IAS 11, which is essentially revenue recognised under long-term contracts and the corresponding cash due, was as follows during 2010 to 2013 as highlighted in the chart below:

Globo Technologies revenue recognised under IAS 11: 2010 to 2013
Source: Globo Technologies annual reports

And finally, adding trade receivables and revenue recognised under IAS 11 together and contrasting this with reported revenue during 2010 to 2013 was as highlighted in the chart below:

Globo Technologies revenue, trade receivables and revenue recognised under IAS 11: 2010 to 2013
Source: Globo Technologies annual reports

GT’s 2013 trade receivables and revenue recognised under IAS 11 is equivalent to 101% of revenue in 2013. This was up from 64% in 2012.

Having established that GT’s revenue was equivalent to 101% of its trade receivables and revenue recognised under IAS 11, now what about the quality of those trade receivables?

Trade receivables were reported as €15.770 million in 2012. Of these, €1.365 million was reported as being above 360 days overdue. Hence in 2012, 9% trade receivables were over a year past due.

By 2013, trade receivables were reported as €12.363 million. Of these, €4.402 million were reported as being above 360 days overdue. Thus in 2013, 36% of trade receivables were over a year past due.

English translation of Globo Technologies ageing of trade receivables
Source: Globo Technologies 2013 annual report

Greek original of Globo Technologies ageing of trade receivables
Source: Globo Technologies 2013 annual report

Percentage of Globo Technologies trade receivables over 360 days past due
Source: Globo Technologies annual report

So we have now established that GT’s trade receivables and revenue recognised under IAS 11 was equivalent to 101% of revenue in 2013 and that 36% of its trade receivables were over a year past due. I reckon this to be an alarming deterioration in the quality of revenue and corresponding cash flows.

Of course in the spirit of openness, it is worth highlighting that:

While 2013 revenue experienced an outright decline of 29% YOY;
and that this (101%) was more than all reflected in trade receivables and revenue recognised under IAS 11;
and that 36% of these trade receivables were more than a year past due;
that GT’s operating profit actually improved (!!!) by 144% to €5.980 million in 2013 from €2.453 million in 2012.

But also worth highlighting is that GT’s net debt increased from €0.780 million in 2012, to €13.731 million in 2013, as illustrated in the chart below, and indicated in the company's balance sheet. This was primarily driven by an increase in long term debt from €0.953 million in 2012 to €15.533 million in 2013. The extra debt was seemingly needed to fund GT's considerable rise in capitalised software related cash outflows.

Globo Technologies net debt
Source: Globo Technologies annual reports

English translation of Globo Technologies 2013 balance sheet
Source: Globo Technologies 2013 annual report

Greek original of Globo Technologies 2013 balance sheet
Source: Globo Technologies 2013 annual report

GT’s cash flow statement highlights that it spent a fair wodge of cash on the purchase of tangible and intangible assets in 2013. GT’s cash flow statement highlights that there was a €12.529 million outflow related to the purchase of tangible and intangible assets in 2013, up from €6.050 million in 2012. The bulk if not all of this appears to have gone on "Δικαιώματα βιομηχανικής ιδιοκτησίας", which is translated as being "Industrial property rights". As this falls in the intangible assets section, I presume this means some sort of software related licensing or development.

So while GT’s net assets were reported to have risen to €25.853 million in 2013, from €22.617 million in 2012, this was more than driven by an increase in reported intangible/software assets to €14.679 million in 2013, from €3.362 million in 2012. I.e., of GT’s 2013 net assets of €25.853 million, 57% (€14.679 million) is in the main, "Industrial property rights" or software related licensing or development.

... Oh and whilst on GT's net book value, do not forget that of €12.363 million in trade receivables in GT’s net assets, 36% are over a year past due, hence 17% of GT's net assets is actually trade receivables over a year old.

... And also recall that software application related revenues fell by €14.480 million in 2013, while seemingly c. €11.047 million in software related expenditure was capitalised in 2013.

English translation of Globo Technologies 2013 cash flow statement
Source: Globo Technologies 2013 annual report

Globo Technologies cash flow statement
Source: Globo Technologies 2013 annual report

Why is this important? It’s important because having been 51% disposed of, seemingly to eradicate unappealing trade receivables, GT is now booked on GBO’s balance sheet with an implied value of €23.7 million. This is almost GT's net book value (NBV), a book value of which 57% (2012: 16%) is comprised of software related development or licensing. A further 17% (2012: 6%) of GT's NBV is attributable to trade receivables which are over a year past due. Further, GBO has marginally increased its valuation on a company where revenue fell by 29% in 2013, and over 100% of that revenue is reflected in trade receivables plus revenue recognised under IAS 11. Further still, while software related revenues plummeted in 2013, software related cash outflows sky-rocketed. In the meantime, GBO's Enterprise and Mobile software business experienced strong revenue growth. In terms of cash flow, GT hasn't generated positive operating cash for two years running. Free cash flow was negative €12.256 million in 2013. GT's net debt has ballooned. And GT's management still owes GBO €9.7 million!

What would Mr Mostafa Khder say about all this?

Mr Mostafa Khder, reviewer of Globo's GO!Enterprise app
Source: Google Play