Iolite Partners commentary for the first half ended June 30, 2016.
For the half-year ended June 30, 2016, Iolite Partners’ unleveraged portfolio returned -14.5% in EUR (-10.4% in USD), net of all fees. Since inception on October 1, 2008, the core portfolio has generated net cumulative returns of +143.6% in EUR (+110.0% in USD) and annualized net returns of 12.2% in EUR (+10.0% in USD). In other words, a euro invested at inception has turned into €2.44 (a dollar invested at inception has turned into $2.10).
This half-year’s performance was highly unsatisfying as measured by market prices and erased all the gains of 2015. Excluding a painful 67% drop of Perion Network (a 29% position as of year-end 2015), the portfolio would have been up in what was a volatile period. I will elaborate on this investment later in this letter.
Good news came in a few days after the end of the half-year, when a large Australian holding reported very strong annual results that immediately lifted the share price by 20% and the portfolio by 6% in a day.
All in all, I consider the portfolio greatly undervalued given the underlying assets and stick with Warren Buffett: “We prefer a lumpy 15% rate of return to a smooth 12% return”.
Thoughts on the Macro Environment
On June 24, 2016, the people of the United Kingdom voted to leave the European Union (“Brexit”). Worldwide, politicians and the media were quick to condemn the voters for a “disastrous decision”.
First, I think it was great the people of the UK had the opportunity to vote on such an important matter. Most other European countries do not put so much trust, confidence and responsibility into the hands of the people (and yes, I am aware the vote was an opportunistic political move by David Cameron).
Second, I think it is presumptuous to condemn the voters. It seems the democratic majority is only welcome for as long as it is supportive of the establishment’s mainstream opinion. Instead of condemning the voters, we should be asking: “What has led to this situation?” and “Is the European Union heading in the right direction?” Personally, I don’t believe the British voted against Europe but rather against the current state of the European Union.
Third, nobody knows what the Brexit will actually look like. Therefore, it is impossible to tell what damage the Brexit will cause to the UK (aside from a possible declaration of independence by Scotland). Switzerland has never been a part of the European Union but has fared rather well being “in” with one leg and “out” with the other. The UK is a much more important strategic and economic partner to EU member states than Switzerland and currently hosts 3 million EU citizens. Therefore, I doubt the EU can play hardball with the UK. It’s also important to remember that the UK has never been a “pure” member of the EU, as it has never been a member of Schengen or the Eurozone.
Fourth, it is likely there will now be similar votes in other countries. I expect far more serious repercussions for Europe if a core Euroland country such as the Netherlands decided to exit the EU and the euro.
Case study: Perion Network (NASDAQ: PERI)
Once we realize that imperfect understanding is the human condition there is no shame in being wrong, only in failing to correct our mistakes. [George Soros]
Fellow value investor Guy Spier once pointedly remarked that there are three types of investment letters:
1) Scaremongering: “The world is a scary place but you can trust me as I have clear vision.”
2) Promotional: “Buy shares of the companies I own [and drive up their prices].”
3) Bedazzling: “Trust me, because I am so smart.”
I would like to introduce a fourth category: “lessons learned”. My goal is not to raise assets by bedazzling a large number of people or to talk my book but to compound my own money – and that of my clients – for a very long time. I believe an open and rigorous analysis of lessons learned is a very valuable tool to help me accomplish this goal.
The financially most painful investment decision in my life to date has been an investment in Perion Network (NASDAQ:PERI). I (and my clients) would be 30% wealthier had I not kept buying the company’s shares over the last few years. This is an investment in what was once Israel’s largest IT start-up, after it had fallen on hard times and looked cheap in the aftermath.
The shares’ steep decline in H1 2016 (-67%) corresponded with the end of a major lock-up period, technical selling after the company was removed from an index, as well as some troubling news from the company itself. Valued at around $2 billion just three years ago, PERI is now trading at an enterprise value of around $100 million. Through a transformative transaction in late 2015, PERI should see revenue growth of 50% as well as solid cash generation in 2016. I believe it could generate $50 million in stable annual free cash flow going forward. Analysts estimate the company is now trading at 2.5 times 2016 earnings.
Unfortunately, there are serious issues with current management: a lack of credibility, a track record of horrendous capital allocation, and an overhead structure that no longer fits the size of the business. Many of these issues only revealed themselves over time and only became obvious earlier this year. It took me too long to spot them and to act accordingly.
In early April 2016, I flew down to the company’s headquarters in Tel Aviv to meet the chairman of the board, management and major shareholders (among them the original founder of the business, an impressive serial entrepreneur who still has a large stake). Looking at the changed nature of the business, I saw a few obvious strategic and structural steps that needed to be taken, but the main goal of my meetings was to remove the CEO (Josef Mandelbaum) and CFO (Yacov Kaufman) to give way to new leadership and a leaner structure.
In the undesirable case Mandelbaum and Kaufman decide to sit it out (and they apparently benefit from the loyal support of a staggered board), I will continue to push for a change as I firmly believe it will require new leadership to realize the company’s potential and to regain the trust of the markets. Sixty percent of the shares are in the hands of five business-savvy shareholders. Their stakes are currently hugely undervalued based on the company’s potential, giving them little incentive to sell and strong incentive to change the way this company is being run.
Investment Thesis (as of June 2015)
I am presenting my investment thesis from June 2015, as this is when I really started to load up on PERI. Most people judge investments and managers based on outcome, leaving them heavily exposed to hindsight bias and outcome bias. In contrast, I believe it is crucial to focus on process and the quality of a decision based on the known facts when it was made. Let’s take a look at thesis vs outcome, and lessons learned.
At a share price of $3.40, PERI is a cheap option with upside >100%
- A challenged and unloved business but a profitable cash gusher trading close to net cash
- Last earnings call: PERI might generate $40-50 million of pre-tax cash flows p.a. with a de-risked search business going forward
- Buybacks (funded from cash, operating cash flows and higher structural leverage once business stabilizes) could materially boost returns to shareholders to well beyond 100%
A Misunderstood Business Model
PERI provides monetization solutions to software developers through a browser extension that gets bundled with their apps and generates income through sponsored search results, primarily from Microsoft (Bing) but also from Google, IACI (Ask) and Yahoo.
This monetization business model has come under increased pressure from customers, Microsoft and Google as some of the industry’s players were overly aggressive in pushing installations and unwanted content. In the case of PERI’s competitor Babylon, a dictionary and translation program was bundled with an aggressive adware program that changed the browser’s home page as well as the default search engine and caused other re-directs as well – and those changes were difficult to reverse. When Google eventually terminated its partnership with Babylon because too many customers complained it negatively impacted their Chrome browser experience, Babylon’s share price collapsed (and took many of its peers down as well).
However, the search monetization business model might not be as bad as those examples suggest. First, search distributors such as PERI materially help nurture the app ecosystem by providing software developers with a convenient and effective way to generate income. Second, search distributors help search engines gain and protect market share. Ultimately, Microsoft and Google need to weigh between the benefits of higher revenues and a strong developer community through search distributors on the one hand, and their platforms possibly losing popularity through overly aggressive monetization partners on the other. Historically, this dynamic has led to interesting cycles in the industry.
Arguably, the industry is currently going through a trough. Over the last two years, Google was extremely hard on Babylon and PERI, and Microsoft cracked down on aggressive traffic generators as well. However, not all search distributors were affected in the same way. For example, it seems Google has remained loyal to and supportive of IACI (Ask).
PERI is the only public search distributor with a strong affiliation to Microsoft. It seems that Microsoft continues to support the monetization industry, albeit on stricter terms. In its Q1 2015 earnings call, PERI’s management commented that “a lot of smaller private companies (were) going out of business”. PERI’s management believed that once the cycle was over, the industry would be smaller overall, with only “three to five relatively major players left in the download monetization side of the search business.”
Indeed, web statistics (e.g. Alexa: www.conduit.com) indicate that PERI’s business is stabilizing at a smaller scale and that it is generating higher-quality traffic than its peers as measured by geographic search origin.
Traffic Acquisition Costs
Historically, PERI would pay app developers an upfront fee per installation of its browser extension. The extension would then generate income for PERI through paid clicks for as long as it remained installed on a user’s PC. In this case, PERI carried the full investment risk, as measured by prepayments to software developers (traffic acquisition costs) vs. cash flow from search income.
To de-risk the business, PERI is switching to a revenue-share model. Under the new model, PERI is just taking a margin of the revenue generated per installation without paying app developers upfront. While the move has an impact on the top line (revenues are lower, as they are now being reported net of TAC), it evens out the bottom line (EBITDA is less volatile as it is no longer swinging with the levels of TAC-spending) and PERI is no longer at risk of having negative returns on traffic acquisition spending. The business will look smaller given the change in accounting, but the quality of reported performance metrics will be significantly higher.
If PERI were to stop all spending on TAC today, it would still generate cash from its installed products base. LTM March 2015 TAC was $140 million (like-for-like). Conservatively assuming an average lifespan of its products of 6 months (guidance used to be 12 months), there is an estimated $70 million in recoverable cash “trapped” in TAC that PERI could free up in a liquidation scenario.
I expect that a smaller but higher-quality monetization business would lift the average lifetime of installations due to lower churn (and thus the recoverable value of reported TAC).
Recent acquisitions are expected to contribute from 2016 onwards
PERI has recently acquired two businesses, GrowMobile and MakeMeReach, that target the increasing complexity of the online and mobile advertising markets. Essentially, they provide platforms for customers to plan, execute and analyze marketing campaigns across various channels (search engines, social media, etc.). PERI’s management believes those businesses build well on the company’s experience of routing and analyzing media traffic.
Both businesses are still very small, but they are growing rapidly. At the moment, PERI is allocating an increasing amount of resources to those opportunities and expensing most of the development costs. MakeMeReach was profitable at the time of acquisition.
At a share price of $3.40, I believe the downside is covered and the upside is > 100%.
At a fully diluted market cap of $250 million (USD 3.40/share), this challenged and unloved but still profitable company was trading close to net cash. The last earnings call suggested PERI might be able to generate $40-$50 million of pre-tax cash flows p.a. with a de-risked search business going forward.
Cash rich: PERI has $116 million cash (= 46% of market cap) and $80 million net cash (= 32% of market cap). In addition, the company reports traffic acquisition costs of $140 million LTM (prepayments/expensed investments into future cash flows).
Improving momentum: management expects PERI’s sales and EBITDA to stabilize in Q2/Q3 2015 and sales to grow from Q4 2015 onwards. Recent acquisitions and investments into online and mobile advertising services (GrowMobile and MakeMeReach) are expected to start contributing to earnings growth from 2016 onwards.
Valuation ($ million)
+ 70: assumed recoverable cash in traffic acquisition costs (TAC)
+300: 6x EBITDA of safer $50 million search business going forward
+ 30: value of GrowMobile and MakeMeReach
+ 20: value of remaining app portfolio – 36: debt
= intrinsic value of 500 ($6.70/share = 100% upside)
- Core business earnings expected to bottom out in Q2-Q3 2015, revenues should grow from Q4 on
- Overall margins should improve as GrowMobile and MakeMeReach ramp up
- Utilization of cash: acquisitions in the monetization space (as the industry consolidates) are likely to generate scaling synergies and thus lead to cost savings and lift revenues/earnings
- Buybacks (funded from cash, operating cash flows and higher structural leverage once business stabilizes) could materially boost returns to shareholders to well beyond 100%
- Take-private: given the business model, PERI might be better off as a private company
- Core business risks: browser changes, erosion of customer base and aggressive competition
- Overestimation of recoverable cash in TAC / negative ROI on TAC (however: in the absence of unexpected policy changes, management should have good visibility based on statistical values and should guide accordingly)
- Value-destructive acquisitions (historically, management’s track record seems mixed at best)
- Dilution through share-based compensation far in excess of value creation over time
01 Dec 15: PERI acquires Undertone for $180 million, mostly paid for in cash and apparently at 6.5x EBITDA. CEO: “The acquisition will be immediately accretive and continues the strategic evolution of Perion into a global technology company delivering high-quality advertising solutions to brands and publishers.”
Guidance for combined business
- 2016 revenue: $350 million (annual growth: 10-15% p.a.)
- 2016 EBITDA margin: 10-12% (15-18% thereafter)
01 Dec 15: PERI receives a $10 million fresh equity investment from J.P. Morgan Asset Management.
31 Dec 15: PERI reveals certain key operating metrics that indicate the search business’ sustainable gross profit might be too low to carry the company’s fixed cost structure.
02 Mar 16: PERI reports underwhelming FY 2015 results, despite a strong earnings boost through the Undertone acquisition (a seasonal advertisement business with a traditionally strong December).
16 Mar 16: PERI provides more disclosure on Undertone and it seems the company is a lot less profitable than management indicated in December 2015.
10 May 16: PERI reports underwhelming Q1 2016 results, despite many accounting shenanigans and despite the benefit of the Undertone acquisition. Management writes off GrowMobile, thus significantly impairing the narrative they’d been spinning for the last two years. The earnings report and the earnings call is extremely light on meaningful disclosure.
Was it a mistake to invest into the company in the first place?
Some people dislike Perion Network because they don’t like the search product. I still believe search is misunderstood, that it adds value to various industry players, and that there is a sustainable and highly profitable niche for the search business – if run by the right management team and on a smaller scale.
In any case, after the Undertone acquisition, PERI has essentially become Undertone – with an attached, smaller search business and a tail of marginal assets.
Why did I underestimate the scale and pace of the decline of the search businesses?
Turnaround stories naturally are subject to earnings volatility and provide limited earnings visibility, so the guidance and narrative given by management matter. While I certainly did my own homework, I put too much trust into what management was saying. Yes, there were warning flags that I rightly spotted. However, I was lulled by an appealing narrative told eloquently by management, the company’s partnership with Microsoft when its peers relied on Google, a very low valuation relative to the company’s strong balance sheet and current cash flow, and business-savvy shareholders with large, illiquid stakes.
While PERI’s quarterly results kept disappointing, something always emerged that kept me hoping for an imminent turnaround. For example, after a particularly dire quarter, the CEO confirmed his bullish outlook and proudly announced he had bought shares in the open market. So I kept buying.
The really bad news only started to trickle in starting December 31, 2015. First, the company revealed certain key metrics that cast doubt on the future profitability of the search business. Second, the company revealed Undertone financials that didn’t support management’s month-old storyline and earnings guidance. Third, PERI had to write-off GrowMobile – thus significantly impairing the narrative Mandelbaum and Kaufman had been spinning. In short, it seemed management had concealed the true (= lower) profitability of search while overstating the earnings potential of recent acquisitions.
Was I right about cash in traffic acquisition cost (TAC)?
Yes. In fact, I believe the majority of the operating cash flow in 2014 ($89 million) and 2015 ($35 million) was generated by liquidating the Conduit/ClientConnect business (reverse-merger in 2014). However, because the ClientConnect acquisition materially diluted existing PERI shareholders (the ClientConnect shareholders ended up owning about 80% of the merged company) and because Mandelbaum decided to bet the house on Undertone, shareholders didn’t benefit.
While I was right about “cash in TAC”, PERI still freed less cash than I had originally assumed. I believe this is because management had compromised on the profitability of customer acquisitions in an attempt to grow the business aggressively (focus on size rather than profitability). I should have paid more attention to the fact management never disclosed any meaningful operational metrics to help me analyze the effectiveness of the company’s marketing spend.
Going forward, I shall pay a lot more attention to the profitability of growth expenditure. I will also put a premium on companies that report detailed metrics to help me measure the return on growth expenditure.
Why didn’t I sell out once I realized what was going on?
PERI always looked too cheap to sell. For most of the time I have been invested in the company, it was trading close to net cash and/or at very low multiples based on current earnings.
There were always plausible reasons to believe the shares’ most recent slide was just temporary in nature: the end of a major post-merger lockup period, margin calls against (former) employees, and forced institutional selling as PERI was removed from an index.
Given PERI’s small daily liquidity, it would have taken a while for me to liquidate my position and I would have driven the price lower. I was watching the large shareholders closely – and found comfort in the fact they weren’t selling either. J.P. Morgan even doubled its stake as late as December 2015 in a capital increase.
Going forward, I shall be more cautious regarding the sizing of my investments.
Why wasn’t the CEO more honest about the company’s situation?
As Warren Buffett’s business partner Charlie Munger says: “I think I’ve been in the top 5% of my age cohort all my life in understanding the power of incentives, and all my life I’ve underestimated it.”
The CEO and CFO are taking home extremely generous pay packages – despite their horrific track record.
Going forward, I shall closely pay attention as to whether bonus arrangements reward adjusted short-term profitability or the creation of sustainable shareholder value as measured in earnings/cash generated per share. These are some of the questions I will consider when considering an investment:
- Did the CEO pay for his stake (or was it granted to him through stock options)?
- What’s the size of his stake relative to his personal wealth?
- Does he make more money by growing the value of his stake or by maintaining/growing his salary?
- Is his compensation based on (adjusted) total earnings or the creation of lasting shareholder value?
- Has he created lasting shareholder value during his tenure?
- Does he see himself as a dealmaker/visionary or as an operational manager?
- Is he focused on size or profitability?
I prefer companies run by owner-managers who make more money by growing the value of their paid-for shares than through bonus payments and stock options.
What were the consequences of the CEO being dishonest (or in denial)?
I believe he failed to cut costs as fast as he should have.
I also believe he intentionally used acquisitions to distract from the issues of the core business and to drive his bonus payments by growing the size of the business – irrespective of the value created to shareholders.
Initially, he went out to buy a few smaller businesses that seemed to play well to PERI’s strengths (data analytics, relationships with app developers). These smaller acquisitions created a nice story to distract somewhat from the decline of the search business and weren’t particularly harmful. In late 2015, when it became clear these acquisitions didn’t work out and Mandelbaum could no longer hide the magnitude of the issues with search, he went “all in” with the acquisition of Undertone.
I prefer a company whose CEO isn’t afraid to face up to a crisis and act appropriately.
Why didn’t the key shareholders step in?
The key shareholders constitute an investment bank (J.P. Morgan), a venture capital fund (Benchmark), and a few business-savvy serial entrepreneurs. All of them have lost vast amounts of money and had their reputation tarnished since they bought into PERI or handed the keys over to Mandelbaum and Kaufman. I can only speculate as to why they haven’t stepped in more decisively but can say this: I didn’t face any resistance when voicing my concerns regarding PERI’s top management and strategic direction. Some seemed to care more than others.
Going forward, I shall talk to key shareholders much earlier. I shall also look closer at the board structure, the quality of the board members, and the personal affiliations of the board members.
Why hasn’t the board fired the CEO yet?
I can only speculate but believe personal loyalties, human complacency, and a lack of deeper understanding are major drivers. Many people enjoy the prestige of being on a board without paying tribute to the responsibilities the position brings. There is no doubt that PERI’s decline has severely damaged the financial standing of its shareholders and (former and current) employees as well as the reputation of Silicon Wadi (Israel’s high-tech industry).
What would I do to take this company forward?
Given the Undertone acquisition, PERI has essentially become Undertone, with an attached smaller search business and a tail of marginal web businesses. I would focus on the two key businesses and right-size the company’s cost structure accordingly. Specifically, these are the changes I’d like to see in PERI:
1) Remove Mandelbaum and Kaufman
2) Eliminate excessive management hierarchies (Group CEO/CFO, Undertone CEO/CFO, etc.)
3) Focus on making Undertone a success
4) Give the head of search leeway and incentivize him to turn search into a stable cash cow
5) Sell or close the marginal web businesses
6) Restructure management compensation (esp. stock options)
7) Introduce and publish meaningful operating metrics to help investors track and analyze the progress of Undertone and search
8) Make reporting more honest and transparent
9) Improve capital allocation
a) Buybacks if possible for as long as the company is cheap
b) Acquisitions that generate shareholder value by growing what is there rather than betting on “the next big thing”
c) Cash dividends if neither a) nor b) make sense at prevailing market valuations
I remain excited about the investment returns I think I can deliver to my partners over the next decades. Virtually all my personal net worth remains invested in the portfolio, as I aim to compound my wealth alongside that of my clients.
As always, I am happy to receive feedback and answer your questions.
See full PDF below.