Stub Position: Long 1 share of CMVT ($6.29); Short 0.13 shares of VRNT ($28.19) --> $2.63 per stub ($1.10 net of cash)
This is an event-driven special situation largely driven by tax issues. This situation has the crucial elements of a successful stub trade: (1) there is significant value to be realized; (2) an event (spin-off plus second-step merger) to unlock the value has already been announced and should act as a catalyst in the next 6-9 months; (3) shareholders of both firms strongly desire the split; and (4) current management has been intent on unlocking the value.
The price action of the stub demonstrates capitulation. Curiously, the last straw was the Jan 11 announcement below. CMVT shareholders were holding out hope that after the accounting issues were dealth with, a PE firm was going to swoop in and buy CMVT (and possibly VRNT as well). This was not impossible, but was going to be messy for the buyer. As well, we do not know what valuation management was requiring for VRNT, either in whole or in part, and I suspect it might have been too high.
But management has clearly taken the next logical step to realize the value: break up the company. And I believe that although complex, they can do so either tax free or with minimal taxation that takes advantage of their ample NOLs. After the break-up, we will have three winds at our back: (1) after a half decade of chaos, I would guess there are some operating improvements that can be made and management will be able to focus on them; I am reminded of Taro Pharma here; (2) as pure plays, the pieces will receive proper value; and (3) each piece will be an ideal takeover target . . . once the IRS allows them to be acquired.
From the January 11, 2012 press release: "Comverse Technology, Inc. (Nasdaq:CMVT) ("CTI"), a global leader in BSS, mobile Internet and value-added services, today announced its intention to distribute 100% of the shares of its wholly-owned subsidiary Comverse, Inc. ("Comverse") to CTI's shareholders on a pro rata basis. CTI is currently exploring, and expects to finalize and announce the structure that will result in the most efficient method of distribution, with the distribution expected to occur in the second half of fiscal 2012. In addition, CTI is exploring alternatives to eliminate its holding company structure either simultaneous with or shortly after the distribution of the Comverse shares."
Although intentionally vague, it is clear that CMVT is planning a spin-off of Comverse plus a second step merger in corporate-tax-speak. The best analysis I have read is found at the Willens Report. I will not violate his copyright by including his report, but the tax analysis I offer below is consistent with his analysis and was arrived at before I had read his write-up.
I am not going to spend a lot of time describing the operations as there have been previous write-ups that address this. But a short summary is below:
CMVT is a holding company with three subsidiaries: (1) Comverse -- 100% owned; (2) Verint -- about 55% owned -- publicly traded under ticker VRNT; (3) Starhome -- about 66.5% owned.
Comverse sells telecom billing and value-added service functions. Think Amdocs.
Verint is unrelated and sells enterprise software, specifically the hot area of business intelligence.
Starhome is much smaller than other two. Its solution is sold to carriers and aids in wireless international roaming. It is the kin of Comverse, not Verint.
There have been previous write-ups that give a better description. The history of this company has been tortuous and a large part of the reason for the stub pricing. The underlying products are undeniably competitive products, but gross mis-management and then the resulting turbulence has left this company battered and bruised.
For this exercise I will assume that the convertible preferred held by CMVT are converted immediately into 10.7mm shares and do not continue to accrete.
mkt cap = (39.3 + 10.7) * 28.19 = $1,410mm
TEV = $1,889mm
Adj Rev '11 = $790mm
EBIT '11 = $135mm (added back all company adjustments except stock comp)
NOPLAT '11 = $100mm
There are some NOLs (PV = $75mm) so it is trading at 18x. The company forecast 10% growth for 2012.
One analysis for VRNT that was found in Barron's a few weeks ago was misleading at best. It stated that VRNT was trading at 10x forward earnings, but this is skewed by: (1) leverage, (2) no taxes now because of NOLs, (3) not including large stock compensation, and (4) funny accounting due to convertible preferred. On an unlevered basis, this company trades in the high teens while growing at 10% top line -- not super cheap as a standalone company.
At this valuation, VRNT as a standalone company is not terribly cheap. However, as a takeover candidate, it may be cheap. Business intelligence software firms like VRNT have been popular targets over the past few years. And VRNT management certainly seems willing to sell. However, as discussed below, one of the tax issues is whether VRNT will be off limits for 2+ years. I believe the gray area is a bit too dark for someone to pursue VRNT until two years after the spin-off/merger. However, if you think that VRNT could be sold without tax implications and that an acquirer would pay a nice premium, then going outright long CMVT may also make sense.
CMVT (sans VRNT)
mkt cap = 207mm * 6.29 = $1,302mm
cash: $481mm less $120mm at VRNT less $6mm minority interest in Starhome = $355mm
ARS: sold in q4 for $49mm
litigation liability: $87mm (on balance sheet at $101mm but paid in q4 with $9mm cash and 12.5mm shares --> $87mm total comp)
net cash: $315mm
VRNT stake: 27.0mm * $28.19 = $761mm
stub TEV = $1,302 - 315 - 761 = $226mm
Note that the stub contains three elements: (1) Comverse, (2) Starhome, and (3) the premium value in convertible preferred over current exchange.
The 2011 financials are mis-leading because they over-state revenue and cash flow is significantly less than earnings as a result of deferred revenue coming off the books. However, management has stated that they expect to be able to stabilize at a true $700mm revenue (i.e. book-to-bill > 1) and 10% operating margins. This is supposed to be conservative as this is how they are targeting their cost structure. Achieving higher revenue would add a lot to bottom line due to high op leverage. Note that Amdocs runs at 15% op margin and trades at 12x unlevered FCF. Assuming the $70mm of EBIT is accurate, we should note that because of almost unlimited NOLs at Comverse, they will not be paying taxes probably ever. So, applying a 12x multiple to $70mm, we get $840mm. As an acquisition candidate, we could expect increased margin with the scale, and $1.2bn may not be crazy.
The book-to-bill will remain less than one for some time and so a 12x multiple is not realistic. However, based on size of deferred revenue ($500mm) and past rates, the bleed should wind down in about 18 months. If so, then a 8-10x multiple today on $70mm FCF is realistic.
The first major risk to this idea is that the business does not stabilize and runs down while producing little cash flow. This is unlikely due to the nature of the product. The business suffered because of its self-inflicted wounds and so new orders stopped as carriers waited to see what happened, but there were no large scale defections. This is a testament to the stickiness of the product. And now, the turmoil has passed. These billing and value-added service solutions for telecoms have very high switching costs. Carriers cannot simply interchange them as they undergo extensive testing before use. As well, the competitiveness of Comverse products has not suffered. Its billing solutions are arguably the best in the industry. There are pockets of growth, but overall, telecom carrier solutions is no longer a high growth area but simply growing with existing customers as new services are offered or existing services are upgraded will provide stable business.
The 2011 numbers indicate $44mm in revenue and $8mm in EBIT. At 65% ownership, this is worth about $50mm to stub value.
So, I would argue that the stub TEV should be north of $600mm rather than the $240mm at which it trades.
There are two clear ways to proceed in terms of separating the company. First, let me discuss the road they are not taking:
1) Spin off VRNT from CMVT. The obvious reason they are not doing this is because if they simply distributed VRNT shares, they would trigger a large tax bill both to CMVT (gains of about $450mm because distributing $761mm of value with $300mm of basis) and to shareholders, who would have dividend tax on the distribution. This is a non-starter.
1 modified) Recap VRNT by giving CMVT 80% of voting power without adjusting the economic share. Then spin-off the VRNT shares. Later, VRNT could exchange the super-voting shares for ordinary shares. Note that both actions (giving CMVT 80% voting power and getting rid of super-voting shares) would require shareholder approval, but in this case, this does not seem troublesome. VRNT shareholders would do it to effect a split. And then the super-voting shareholders would do it for liquidity. This may seem contrived, but the IRS has OK'ed such a move in the Metlife-RGA spin-off. See the Metlife transaction in the attached article:
I do not understand why they do not go this way, and the only reason I can think of is that they want the NOLs sitting at CMVT to go to VRNT, not Comverse. However, if the IRS denies tax-free status to the proposed spin-off plus merger, then I would expect them to pursue this route.
So, the path that they are pursuing is two steps:
- first, spin off 100% owned Comverse from CMVT.
- This alone is not tax free since CMVT is a holding company. So, effectively simultaneously, they will merge CMVT and VRNT. Note that because VRNT owns greater than 50% of combined CMVT/VRNT entity, the Morris Trust rules are satisfied. As well, the merged CMVT/VRNT is clearly a bona fide operating company so that now the spin-off of Comverse is tax free. And all NOLs are preserved and use-able.
The problem with this method is that the Starhome stake is now at the new VRNT, not where it should be, which is Comverse. This is a small inconvenience and could be corrected with a small tax bill. For example, Comverse could simply purchase Starhome stake from newVRNT for $50mm and VRNT would pay corporate taxes on this $50mm less the basis. So, this is a $15mm inconvenience.
I do not think this is a terribly controversial structure, and I believe the IRS will approve. On the positive side, the two companies have absolutely no business reason to be together, and the two have operated as separate entities for a long time.
The next issue is: When can Comverse or newVRNT be acquired without triggering a retro-active tax event? This is tricky and we do not have enough information to answer. This is governed by Section 355(e).
Although not strictly determinitive, in most cases, 2 years after spin off forms a safe harbor. The question here is whether these companies could enter into discussions as soon as six months after the spin-off/merger. There can be no "plan" and there must be a "fit and focus" purpose to the spin-off/merger. In this case, the "fit and focus" test is fairly easily passed. However, we do not know the exact nature of discussions that CMVT had with potential acquirers and whether this removes the possibility of a quick sale or even eliminates the 2-year safe harbor as well (this is very unusual for a publicly traded company using credible bankers).
Insiders have more information on this point because they know the exact nature of discussions that CMVT had over the past few quarters.
I am not counting on a favorable outcome of this issue, and I am valuing based on both entities remaining standalone entities, but it is a nice possibility and could drive a premium in both equity prices.
There are two core risks: (1) My tax analysis is flawed and there is no manner to split Comverse from VRNT that is not too costly; (2) Comverse business is less stable than I believe and slides badly.
In Q2 or early Q3, I expect the company to finalize the restructuring plans after receiving feedback from IRS.
In Q3 or Q4, the actual spin-off plus merger (or other restructuring) will occur.
In 2013 or 2014, both entities become acquisition candidates.
|Subject||RE: straight long?|
|Entry||02/07/2012 02:28 PM|
I address that in the write-up. I do not love VRNT as a standlone long at this valuation, but if market thinks they can become acquired six months after spin-off/merger, then I think spin-off/merger will provide catalyst for VRNT share price. And if they really can sell themselves without tax consequences six months after restructuring, then I do like it quite a bit as an event-driven long.
|Subject||RE: couple questions|
|Entry||02/08/2012 08:58 AM|
I think all three of your points are valid:
1) In terms of return, I agree this is all about the denominator. The most extreme version has $2 of upside on a $1.10 stub net of cash. The most extreme version the other direction has $2 of upside on a combined long-short position of $10 (long $6.28 stock, short 3.70 of VRNT stock). This is all about how your fund treats risk and capital allocation. I argue that the risk-reward here is pretty good when all things are considered. But if your fund would consider this a 20% return on capital even is fully successful, then I agree it is mediocre.
I should note that I rarely do stub trades because I feel there is no catalyst and you end up with this blob on your books that will never converge. I do think this situation is different.
2) On the 10% operating margins, I admit I was a little casual. I am very familiar with this industry and their products and there is absolutely no reason why they cannot achieve 10% operating margins in a stabilized business on their core business. The past results have never come in a stable environment for a long time. I do think anyone looking at CMVT needs to get comfortable with this by talking to folks in the industry about their product and looking at their past mis-management.
One way to value a moat is to consider how bad you can behave and still retain customers. It is striking that Comverse still has customers.
3) Your next point is about how equitable the separation terms will be. This is a great question, and I was going to get into it. One thing to do here is to slightly underhedge the VRNT. But I do not think necessary because while CMVT has no other alternatives, neither does VRNT. This is why I pointed out that both shareholder bases badly want a separation. As such, I think that CMVTstub will get fair value for its shares and tax assets that do pass to VRNT.
It should be noted that VRNT shareholders by their own belief have much more to gain from the separation.
|Entry||02/08/2012 11:46 AM|
Why do you assume the preferred is converted into common instead of using the (higher) liquidation preference, which will continue to accrue at 3.875%?
|Entry||02/08/2012 12:07 PM|
Fizz argued I was too generous to CMVTstub in my write-up, and you are implying the opposite! One of you is correct. I believe it is you.
I was just being conservative. Your point is a counter-point to fizz's question about whether CMVTstub is going to get screwed by VRNT in the separation. The worst case seems to be that CMVTstub loses its tax assets (and gets nothing) and only gets 10.7mm shrs for pfd.
We can argue different scenarios going the other direction. My valuation has what I think is a conservative but not worst case scenario.
You are arguing that I am being too conservative, and you may be right. CMVT should get more than 10.7mm shares for pfd both because of liquidation preference and because of future accretion (while VRNT shares are not paying dividend), even if converted value exceeded liquidation preference. Tough to quantify but I agree with your sentiment. I would argue this probably adds 25 cents (ish) to stub value. I pointed out that pfd had premium value over 10.7mm shares but left out valuation because too vague, and I wanted to focus on the tax issues, which I believe are the most important here.
|Subject||RE: stock action|
|Entry||03/19/2012 12:43 PM|
I do not have any insights into the trading of the stub or the stock outright. Given the nature of the stub, the stub is amazingly volatile on zero information. Most of the volatility is CMVT stock (as opposed to VRNT stock). Look at today. It trades thin. And clearly lots of exhaustion here.
|Entry||08/03/2012 09:26 PM|
I am not sure if anyone follows this (based on rating, no one does; and based on performance, you would have been correct), but . . .
Starhome sold for just where I had estimated. 65% stake sold for $52mm. This helps in that Starhome did not belong with Comverse business, and so I imagine this was plan all along.
So now Comverse spin-off pure-play and we await the announcement and details of the CMVT/VRNT merger.
The performance of comverse piece has been poor, but I continue to like the stub idea.
|Subject||RE: Starhome Sale|
|Entry||08/03/2012 09:44 PM|
Upon closer inspection (note to self: do not simply read forbes blog), they received significantly less than $50mm I had projected. They only received $37mm, and part of this is in escrow for 18-24 months.
I do not understand how they received so little. It does appear they gave $5mm to other shareholders to get the deal done (clearly, comverse wanted deal done -- which makes sense). But this only comes to $42mm and so not sure where remaining $10mm went -- seems like a steep fee even for i-bankers.
So this knocks off 6 cents per stub share in value.
|Subject||RE: Deferred revenue and cash|
|Entry||08/23/2012 10:21 PM|
You are thinking about the mechanics of deferred revenue and related deferred cost of revenue properly. However, I think the specifics with respect to CMVT are a bit more complicated than your numbers. I must state upfront that I have not been able to get 100% clear answers.
1) The margin of satisfying the deferred revenue obligations may have a much higher margin than you are assuming. In other words, the heavy lifting is done early in the contract period while revenue is recognized more evenly. I have never gotten an answer to this directly, but my guess is that it is partly true.
2) The deferred revenue is not going to go to zero even with their new revenue model as you assume. The normalized revenue model will have deferred revenue.
3) There also appears to be some offsetting move in prepaid expenses and not only deferred cost of revenue, and possibly even AR/AP working capital.
4) Mgmt made two statements on the June call that are relevant: First, they stated that they expected the revenue model to normalize in 2014: "We will get to that normalized state, will it be 2013? I think our modeling shows 2014. We have got $540 million of deferred revenue, we've got backlog of $982 million, we're working that down, so you will see a book-to-bill less than one, and revenue in excess of the underlying bookings and maintenance stream for 2012 and 2013, and I would say in 2014, we would get to your more normalized state." Second, they expect 2012 to be "modestly cash flow positive."
If both these statements are true, then they seem to be saying that the hit to cash is closer to half your $181mm estimate.
The conclusion that I came to in analyzing their financials was that they will be cash flow neutral for 2012 and 2013 before normalizing and then producing their $700mm revenue, >10% op margin repeatable biz without working cash flow fluctuations.
I have a question for you.
I noticed that as of 1-31-12, the total cash (including restricted) in the consolidated entity was $558mm, and the cash at Verint was $164mm. But on the Form 10, the pro forma number was only $190mm. No big deal, I thought, as the other $200mm must still be at CTI parent co. And mgmt verifies the number explicitly in their June call. However, in the latest press release regarding the merger, I do not see any mention of it. Further, the press release announcing the merger states that after spin-off, CTI has no assets other than Verint stock. I cannot believe they would OK a merger where they gave up the $200mm to Verint. I have contacted IR but have not heard back. Is there some obvious way to reconcile the cash balances? Where is the missing $200mm?
|Subject||RE: RE: Deferred revenue and cash|
|Entry||08/24/2012 10:45 AM|
I spoke to IR and received an answer regarding the missing $200mm of cash. The original Form 10 does assume that $200mm of cash will remain with CTI. However, this will be changed and all cash balances (after paying admin fees, etc.) will be pushed down to Comverse (spinco) for the spinoff.
So I think the updated numbers are:
CMVT shares at 6.09; VRNT shares at 28.95.
CMVT mkt cap is 219mm * 6.09 --> $1,334mm (note the increase in shrs due to litigation settlement)
They will get 28.4mm shares of Verint, including the incentive/premium payment --> $822mm
They received $37mm for their share of Starhome.
After fees and such, I estimate about $325mm of total cash -- both restricted and unrestricted.
I should note that all the restricted cash becomes unrestricted within 24 months.
So, TEV of spinco is $1334 - 822 - 37 - 325 = $150mm.
If I am correct about the deferred revenue costing about $100mm, which I do believe is consistent with their statements, then we have adjusted TEV of $250mm. And for this we get the $700mm/10+% margins (less corp overhead) business that mgmt promises.
There are some nips and tucks from the original write-up (we got less for starhome; but more for Verint; etc.; stock prices have gone against), but it is essentially the same valuation. I would say that the results have been a clear negative, but mgmt has steadfastly stuck to its forecast.
If you believe mgmt, this is cheap. I am less confident given the results thus far, but I am still a believer.
|Subject||RE: Author Exit Recommendation|
|Entry||12/30/2013 02:47 PM|
I entered this stub position almost 2 years ago at equivalent to buying CNSI at $26.30.
The stock is up 50% (and TEV almost doubling) over that time period, but the truth is that I do not think the operations information flow has been positive. Most of the rise has happened in the past month, presumably due to presence of activist investor.
I do not think CNSI is going to get to cash revenue of $600mm and op margin (even if we do not include stock comp and amort of intangibles) of 15% by the end of next year. And as a standalone company, I do not think cheap at this price.
I do think a strategic acquisition makes sense at a healthy premium ($50) due to significant synergies, but otherwise, paying too much. To hold on here, you must be optimistic that company sells self in 2014.