Top Image Systems is a Nasdaq-listed, Israeli domiciled document capture software provider.  Its software is used to capture information from paper and electronic records and feed it into accounting, content management and other systems.  The software is commonly used to capture documents like invoices, census forms and mortgage applications.  Revenue comes from license fees paid upon installation (50%), engineering fees for customizing the software (25%) and maintenance fees (25%).  The majority of their sales are in Europe (c. 75%).  

Sales grew 32% in 2011 and management have guided 17-23% in revenue growth for 2012, with operating income of c. $4.4 million.  My primary research indicates that this is a reasonable target and revenue should continue to grow at this rate for several years, with profit growth in excess of 20%.  The company has $18.5 million in NOL carryforwards so it will not pay tax for 2-3 years.  It is trading at about 10.7 times forward earnings which I think is far too cheap for a business with very high return on tangible capital that should enjoy operating profit growth of over 20% for at least the next few years.  Recent M&A activity in Top Image Systems' industry has priced comparable businesses in the region of 2.5 times sales (vs TISA's 1.6x trailing EV/sales multiple or 1.4x forward EV/sales multiple).

The prospects of the business have improved dramatically over the last three years due to an efficiency drive led by the Chairman and CEO after they acquired 1.7 million shares from Charterhouse in February 2009, a time when the company was close to being seized by debenture holders.  They hired a new CFO, discontinued unprofitable sales of 3rd party products, shut down unprofitable offices and fired 168 employees (mainly in the Far East).  This saved the business and allowed them to pay off the debenture in full by 4th October 2011.

I think Top Image Systems is worth at least 60% more than the current share price and is undervalued for a number of reasons:

• Until the debenture was paid off in Oct 2011, TiS was paying very high interest ($928k during 2011 or 27% of operating income, $2.2m in 2010 or > 100% of 2010 operating income).  This depressed net income so screens for fast-growing cheap stocks would have excluded the company.  Now that the debenture has been paid off in full the interest burden will be nil going forward.

• The sustainability of the company’s competitive advantage is poorly understood.  Potential investors focus on TISA’s intellectual property, which is not a strong barrier to entry but my research calls indicate that the bulk of competitive advantage comes from customer switching costs (see barriers to entry discussion below).

Business overview:
Top image systems is a niche player in the document capture industry.  Total industry revenue in 2011 was about $2.4 billion, with the top 10 players taking 58% of that pie.  These players provide the 'onramp' for getting paper and electronic documents into the accounting, workflow and document management systems of customers.  Competitors include Kofax (17% market share), Readsoft (11% market share) and EMC/Captiva (7%) market share, although each of these focuses on lower-volume and less customized installations than the typical Top Image Systems contract.  Application examples include capture and insertion of invoice data from paper and pdf documents into SAP/Oracle, data capture from mortgage forms and census form capture.  

Top Image Systems specializes in high-volume installations (millions of scanned documents per year) and has established a strong niche in this area.  The company's solutions combine multiple OCR engines and use a proprietary voting system to achieve higher recognition rates than competing software.  The result is a more expensive but more accurate solution.  The difference between 94% recognition rates in a competing system and 97% automatic recognition in a TISA system can mean large labor savings for a client that processes millions of documents per year.  

Examples of high-volume applications that TISA sells include:

• Census form capture.  TISA has won contracts to capture data from German, Italian, South African and Argentinian census forms in the last year.

• Centralized 'Mailroom' installations. This is where a single system scans for the entire organization and automatically distributes documents to the desired departments and software destinations after the OCR process.  A recent example was a contract worth $1million in revenue to provide BARMER GEK (a German state health insurance provider) with a mailroom system.

• Customized high-volume capture products for financial services firms.  A recent contract was to provide a customized system to the Asian-Pacific arm of Bank of America.  Past customers include Standard Chartered and Citi (past installations were on non-US sites).

Conversations with industry salespeople indicate that switching costs are quite high both for the end customer and for channel partners (40% of TISA sales).  Past customers wishing to introduce the latest developments in capture software to their organization will typically find that it much less expensive to upgrade the systems they installed in the past to the latest version from the same software provider rather than switch to the solution of a competing provider.  Reasons (listed in descending order of importance):

• Higher risk of business disruption involved in switching to a completely new system.  The business-critical nature of the documents being ingested make this a significant consideration in many cases.

• Higher costs of IT expenditure that would be required to take out the existing system and integrate a new system into the existing infrastructure (accounting system, content depositories etc).

• Familiarity with and acceptance of the existing interface by users.

• Hardware changes if existing scanning equipment and servers are not compatible with the new software.  

The cost for channel partners to switch to selling software from a different provider is significant for different reasons:

• There are significant training related costs required before the existing salesforce/installation crews become fully productive.  It can take 1-2 years before personnel fully understand the technology and how to avoid installation pitfalls to get the highest recognition rates in a given environment.

• Channel partners use their existing customers as references to win new business.  If they switch to using competing software, they will find it difficult to win new business because they will not be able to demonstrate real-world working solutions using that software.

• Channel partners often enjoy maintenance revenue from their existing install base.  If the channel partner tries to add a new capture providers’ software to its product portfolio, they will have to pay support staff with expertise in both systems to keep this maintenance revenue stream flowing.  The high fixed cost of maintaining this staff is a barrier to switching for many of these small firms.

Channel partners’ hesitance to switch software providers can be seen from Kofax’s results since they implemented a 'hybrid' sales model in 2007, selling directly for high-value contracts and through the channel for smaller contracts.  This was perceived by partners in the channel as an attempt to sell directly to their most lucrative customers, which created a lot of badwill towards the company.  However, due to the barriers to switching discussed above, few partners dropped Kofax’s products.  Typically the worst action taken was to add a second product to the channel partner’s suite in addition to Kofax.  Therefore Kofax’s channel business was not destroyed like one might have expected would happen in a business with lower switching costs.  My estimate is that Kofax lost about 15% of their channel sales as a result of this move, which is less than I would have expected considering the negative reaction from partners that I have heard about through scuttlebutt.

Fixed R&D costs present a less formidable barrier to entry, but nonetheless may keep smaller entrants from coming to the field.  The businesses in this industry typically reinvest 6 – 15% of their annual revenues in R&D.  Since Kofax’s revenues are much higher than Top Image Systems’ or Readsoft’s, one would expect their higher aggregate R&D spend in dollar terms to negatively impact both of those businesses.  However TISA and Readsoft found a niche to specialise in and were able to match Kofax’s research output (and market share) in that niche.  Top Image Systems focused on high-volume solutions and Readsoft focused on the accounts payable processing segment.  

The combined result of these barriers is high returns on tangible capital and stable market shares for the companies in the industry that are run reasonably well (Top Image Systems in the period since the efficiency drive mentioned above, Readsoft), and continued existence even for businesses that are run poorly.  Readsoft's return on tangible capital was over 60% in 2011 and TISA's was over 100%.  Readsoft's market share in 2004 - 2010 ranged from 12 to 15% while Kofax's grew (partly due to acquisitions) from 6% in 2006 to 17% in 2010.  Top Image Systems has maintained market share at just over 1% since the reorganisation in February 2009 but has a much larger share in the high-volume niche (market share statistics are not published for the high-volume segment, my opinion here is based on analysis of contract win announcements and discussions with industry sales executives).


Growth potential, addressable market, potential European sovereign crisis impact:
Figures published by Harvey Spencer indicate that historical growth in the worldwide capture market over the last 5 years has been in the low teens range, with industry sales growing at just below 10% even in 2009  as cost-cutting businesses adopted the industry's labor-saving software in an effort to adapt to the recessionary environment.  Barring a complete financial catastrophe, I think it is reasonable to expect at least high single digit to low teens growth for the industry as a whole.  Companies’ desire for greater efficiency has not abated. Penetration of the industry’s products is relatively low and ROIs for customers continue to be very high.  A report from the Hackett Group indicated that only 17% of Fortune 1000 corporations use invoice automation and extraction as their primary solution for handling paper invoices .  70% handled invoices either by scanning with manual data capture from the scanned image (no OCR) or by manual capture of data from paper invoices.  Therefore the remaining addressable market is sizable.  Return on investment from installation of document automation software can be north of 100% in 12-18 months due to savings related to manual data entry, systematic capturing of early payment discounts and reduced administrative errors.

Top Image Systems’ sales have been growing much faster than the industry, with sales in 2011 up 32% YoY.  A few factors have been driving this sales increase, some of which can be expected to continue:

• A trend has emerged where customers seek a single solution for processing documents for many departments rather than each department buying their own solution.  These ‘mailroom’ type solutions are typically higher-volume (because all the documents for the enterprise are handled centrally).  This is an area of focus for Top Image Systems and it has announced 38 large mailroom wins to date (typical contract size $500k - $1mm).  Calls with competing sales force executives and analysis of corporate news releases from the industry indicate that this trend towards mailroom solutions should continue.  

• The second area of focus for Top Image Systems, its banking platform, has begun to gain traction, with contract wins in the Asian-Pacific arms of European and U.S. banks.  The initial contracts will make future sales pitches to banking customers easier (because they have a stable of existing customers for the same solutions which they can use as references) so growth in this area should continue.  During 2012 the company plans to open a direct sales US sales office in an effort to drive US sales of banking products.  For the industry as a whole, over 50% of sales come from North America .  With less than 15% of sales from North America, Top Image Systems is highly under-represented in the U.S.  Any traction on correcting this imbalance should result in faster than industry sales growth.

• Competitor Kofax (17% market share) has been undergoing significant reorganisation of its route to market over the last few years, with large reorganisation expenses (read: redundancy payments to former salespeople) being paid each year.  Top Image Systems and others have been able to hire some of Kofax’s former channel salespeople as a result.  In the Q4 conference call, the Chairman noted that 11 salespeople were added during 2011.  It was noted in the prepared remarks of the Q3 conference call that channel sales have expanded significantly during 2011, with 7 new channel partners from the EMEA region added to their customer list.  The hiring of Kofax salespeople will continue to boost sales during 2012 because there is a 1 to 2 year lag between when a salesperson is hired and when they become fully productive.

In the Q3 2011 conference call, Chariman Izhak Nakar said that the pipeline of deals that they are currently tendering for was at about $100 million, up from $70-$80 a year prior.  If TISA can close these prospective deals at the same rate that they did during 2011 (38%), sales would grow by over 30% this year.  Nakar indicated in the Q3 2011 conference call that barring the unexpected, “we don't see any reason not to think that the next year [closing rate] will be different from this year.”  A look back at the historical conversion rate of announced pipeline contracts to subsequent year sales shows that the closing rate was 38% in 2008, 28% in 2009, unknown in 2010 and 38% in 2011.    

A messy outcome to the sovereign debt / banking problems in Europe (c. 75% of TISA sales) would obviously make sales come in below this level but should not be disastrous for the firm judging by the results from 2009.  In 2009 Top Image Systems sales dropped by 27% ($8.7 million) YoY, however $7.5 million of the drop was due to the discontinuation of low-margin 3rd party hardware and software business which is no longer part of the business.  The 2009 revenue drop from products that TISA continues to sell today was only about 5%.  Competitors Readsoft and Kofax both saw sales grow throughout the recession.  The following factors moderate the impact of recessions on the TISA's sales:

1. The company's products are used to increase operational efficiency and is therefore attractive to companies looking to cut costs in a recession.
2. 25% of sales are from recurring  maintenance expenditure.
3. 25% of sales are from service revenue which extends about 1 year from the initial sale.
4. About 7.5% of license sales are to governments capturing census forms.  Since census exercises are scheduled on a regular 5 or 10-year basis, this revenue is generally not impacted by recessions.

Management, capital allocation, operational review:
The current market value of shares owned by CEO Ido Schechter is over 5 times his annual compensation and over 15 times Chairman Izhak Nakar’s annual compensation .  Although net worth data is not available for either Schechter or Nakar, I suspect based on a review of their career history that their shareholdings in Top Image Systems constitute the majority of their net worth.  

A major blemish on the capital allocation record of both Nakar and Schechter was the purchase of Asiasoft Global Pte during 2007, using capital raised in a convertible debt offer a few months prior to the deal.  While the purchase price looked okay on a [Price / Revenue] basis at $3.8 million for a business with $4 million in revenue, it quickly emerged that Asiasoft was not going to be profitable in its existing form and within 18 months the majority of its workforce had been let go in a restructuring of operations.  TiS took a $1.84 million writedown on the goodwill that resulted from the acquisition.  The Chinese, Hong Kong and Malaysian offices of Asiasoft were shut down after the acquisition, with only the Singapore office remaining open.  

However, incentives for Nakar and Schechter were very different in 2007 when the Asiasoft acquisition was made.  Back then Nakar only owned 3.8% of the shares outstanding (he had not yet bought the Charterhouse stake), worth just over $1 million at the time.  Schechter’s ownership was not even large enough to merit a mention in the company’s 20-F.  Nakar’s contract entitled him to receive one-time payments of $100,000 to $200,000 for every acquisition made (presumably because a major part of his job was to find suitable acquisition targets).  His cash salary from TiS in 2007 was only $160,000.  Therefore he was heavily incentivised to make acquisitions, even at prices that were likely to result in a share price drop of up to 20%, because the subsequent financial pain would be offset by the acquisition closing bonus.  This is not the case today because Nakar owns a far larger stake in the business, worth over $7 million at current prices.  His pay package no longer includes a bonus for acquisitions.  Instead he now gets paid remuneration based on the number of days he spends working in the business.  Schechter now owns a stake worth $2.8 million.

If his recent pronouncements are to be believed, the near-death experience for the company seems to have prompted Nakar to find capital allocation religion.  In a recent news article he commented: “I have no reason at present to carry out any acquisitions," admitting that the failed acquisitions that were carried out with the monies from the debenture prevent him from taking a similar action. "The only thing that the debenture issue gave the company was a big headache. "  In addition, he has commented publicly on how undervalued he thinks TISA's shares are.  

Combining the new incentive structure for management with these statements, I think it is reasonable to conclude that capital allocation will not destroy value going forward.

Income statement data for the last 5 years and my 2012 estimates with notes on each line item are here:  http://www.scribd.com/doc/85395296?secret_password=94iph2h620pb8sd00fl

It should be noted that prior to Izhak Nakar's return to being active full-time in the business in Feb 2009 and the hiring of the current CFO shortly afterwards, Ido Schechter and the previous CFO were not doing a great job of running the business for profit.  However in February 2009, after the recession had hit and the extent of the Asiasoft acquisition disaster came to light, management bought a significant amount of stock as mentioned above.  Since then they and the new CFO Gili Shalita have led an impressive return to profitability at the company, with operating margins rising from negative levels to 12% and return on tangible capital rising from negative to in excess of 100%.  The change occured due to discontinuation of low-margin 3rd party product sales, the closure of unprofitable offices and rationalization of the workforce, with 168 employees let go (mainly in the Far East).  Based on the new incentive structure for management and comments from the company on the continued profit-oriented focus of the business, I believe that the recent strong financial performance will continue.  As a comparison, operating margins at Readsoft and Kofax are within 1% of TISA's current levels.  Management at both Readsoft and Kofax have commented that they see margins going to c. 20% in the next 2-3 years vs their c. 11-12% levels today.

Conclusion, valuation:
For 10.7 times prospective 2012 net income, investors in TISA shares are getting a business with no debt, probable 20% sales growth and greater than 20% operating income growth for the next few years, very high returns on tangible capital and a management team whose incentives are well aligned with shareholders.  Even if this growth tails off after 2 or 3 years, I think the stock is worth at least 60% more than what it is selling for today and has the potential to be worth far more than that.

Recent M&A activity in the industry has priced similar businesses at over 2 times sales.  The most comparable transaction was the recent acquisition of Brainware by Lexmark.  Scuttlebutt indicates that Brainware's revenue was c. $60 million and Lexmark press releases show it was bought for $148 million.  One can argue about whether or not this multiple made sense, but the difference between the implied value of Top Image Systems using a similar multiple (c$6 per share) and the current share price ($3.85) is large enough that even if Lexmark overpaid, TISA is still undervalued by comparison.

-- Market realization that net income will be substantially higher now that the debenture has been paid off in full
-- A banking platform contract win with the US or European arm of one of the banks which TISA has done business with in APAC.
-- Continued sales and operating income growth showing up in the financial statements