In a report titled “Powering Up the Sector” released today by the Susquehanna Financial Group, analysts Jason Kraft and Chris Kwak look at the stocks of several leading publically traded video game companies, and discuss why they have upgraded these stocks from 'Neutral' to 'Positive', giving plenty of insight into the game biz through 2010 in the process.
The report starts as follows: “We are upgrading the shares of ERTS (EA), ATVI (Activision), THQI (THQ), and TTWO (Take-Two). We believe estimate revisions have sufficiently accounted for risk – namely, lower peak earnings scenarios. We also believe what to date had been underappreciated downside scenarios have quickly been absorbed by investors, and recent pullbacks in stocks (while some are company specific) point to a capitulation. Our view is that sentimental factors are now dragging stocks below fundamental support levels – that is, our peak earnings forecasts have not changed much, and given the multiples at which the publishers are trading, the risk-reward indicates modest downside and meaningful upside from current levels.
We think simply about the video game publishers: What drives these stocks? Momentum and upward revisions of earnings estimates. When do we see this? Heading into a new console cycle. The problem has been one of timing. When is it too early to play the next-gen cycle? When earnings revisions are downward. Finding that inflection point where both sentiment and fundamentals are bottoming has been the billion dollar question. Have we reached the bottom in earnings revisions? We have. But if we are wrong, we believe it won’t be by much.”
“In the “When The Levee Breaks” issue of our Video Game Journal on March 7, 2006, we highlighted downside scenarios for publishers. We asked investors to consider the following: 1) PS3 delay; 2) current-gen unit sales decline faster than expected; and 3) current-gen pricing erodes further. We argued that these scenarios may uncover a publisher’s potential revenue and earnings trough. Determining the probability of each scenario was difficult (and beyond the scope of the report), but we encouraged investors to contemplate what might happen if the levees were to break, and the disaster scenario came to life. Accepting this possibility would be painful, but it would represent the trough for the industry and the stocks. We argued the sooner we accepted these scenarios as probable – and not just possible – the sooner we would reach the bottom.
In the next issue of our Video Game Journal, “What Could Peak Earnings Look Like?” (“Peak EPS”) on March 21, 2006, we published our peak earnings forecasts for the publishers. We knew how bad things could get, but how good could it get, and when? We noted that arriving at peak earnings sets the stage for valuation. In that issue, we did not discuss valuation. We do here.
The following scenarios do not represent our published estimates. They go beyond our published forecasts. They are, nevertheless, our best guess.
We project revenue from CY06 through CY10, acknowledging that CY10 may not be the last year of the cycle. We categorize revenue into buckets (console, portable, PC, et al). Each revenue bucket is of differing size and growth rate; each revenue bucket boasts margins at differing levels and rates of expansion. Traditional buckets include console, portable, PC, and distribution (if any). Newer buckets include wireless/mobile, in-game advertising, and micro-transactions.
– Consoles. If we assume the hardware installed base grows 10%, and average software prices (ASP) increases 20% across the board ($50 to $60 for AAA titles), this implies software revenue could grow around 30% (holding attach rates constant). [We argued in the “Attach Rates” issue of our Video Game Journal on June 6, 2006 that attach rates could fall cycle-over-cycle.]
– Portables. We believe portable software revenue could grow over 140%. The portable installed base grows more rapidly since the market now includes two major Nintendo platforms (GBA and DS) and Sony, and average pricing grows because higher-priced PSP titles elevate the industry’s average. PSP sales could take a hit when PS3 and Nintendo Wii launch.
– PC. PC revenue growth may be a function of 1) the burgeoning popularity of multiplayer online games; 2) Microsoft Vista replacements; and 3) the already large global installed base.
– Wireless, In-game Advertising, Micro-transactions. We assume wireless is the fastest growing “new” revenue bucket due to the large global mobile phone installed base, while in-game ads and micro-transactions may need more time to develop as markets.
Within costs, we assume separate cost of goods sold for different platforms and different components, including product costs, software amortization (where appropriate), and royalties. Within the traditional revenue buckets, we believe PC carries the highest margins (lowest costs), followed by portables, next-gen consoles, and distribution. We believe the new revenue buckets carry very favorable margins; however, their lower estimated revenue contribution limits their impact on the P&L. Operating expenses are determined by two factors: headcount and advertising.
– Gross Margin. We estimate the following gross margin ranges: consoles 35-60%; portables 55-70%; PC 80-90%; wireless 65-75%; in-game ads 80-90%; and micro-transactions 80-90%. Distribution gross margins range between breakeven and low double-digits, depending on the publisher’s hardware/software revenue mix.
– Operating Margin. Headcount and advertising metrics drive our operating margin forecasts. We expect operating margins to be the lowest for consoles given larger team sizes for next-gen development. Again, portables, PC, and new revenue buckets generally have lower operating expenses, and these revenue buckets need to grow rapidly to offset likely operating margin erosion from next-gen consoles.
Our analysis shows that peak revenue and earnings for the four publishers typically occur in the last or second-to-last year of the next-gen cycle. In the next-gen cycle, we expect none of the publishers will have peak operating margins that exceed those in the current-gen cycle. This reflects our assumption of higher software development and licensing costs for next-gen titles. Despite the industry’s current pain, the industry’s long-term outlook could be positive.
The upcoming cycle will probably be longer than the one we’re exiting for many reasons. We believe the distance from trough to crest (peak EPS) will be smaller. However, we think the cycle will be elongated, and as a result, it may be possible for publishers to sustain peak earnings power longer.”
“What are these stocks worth? Applying reasonable multiples to the four stocks, and discounting them to next-year’s calendar year end, we arrive at the following valuations. These prices target the end of CY07. We remind investors that in the February 7, 2006 issue of our Video Game Journal, we arrived at an asset value of $10-$13 per share for TTWO’s Grand Theft Auto
franchise. Given the company’s $2 in cash, a $12-$15 range should provide a floor to the stock.
We project cumulative cycle-over-cycle revenue growth of 58% for Electronic Arts, driven by strength in consoles and portables. EA’s annualized (and popular) sports and racing franchises could drive 34% growth for consoles, in excess of our baseline 32% estimate for industry consoles.
The two aforementioned genres are very large and offer EA ample opportunity to exceed the industry baseline. We believe consoles will be the largest bucket, but it may yield share to other platforms. We believe portables will demonstrate the highest growth, driven by the PSP. Currently, EA publishes 21 PSP titles and will most likely expand given the company’s large portfolio. We expect PC revenue to grow 27% cycle-over-cycle. We believe The Sims
franchises can sustain that level of growth. We believe newer buckets such as wireless (via JAMDAT acquisition) could drive total revenue growth in the last year of the cycle (peak total revenue of $4.8 bln in CY10), when console revenue typically falls in anticipation of a new hardware platform. In the current-gen cycle, growth in the sixth year (CY05) was hurt by a sharp downturn in current-gen demand.
EA does not disclose cost of goods sold components in dollar terms. As a result, we estimate blended, corporate gross margins under the assumption that the company’s higher anticipated royalty expenses will temper gross margin expansion. As of December 31, 2005, EA had $1.43 bln in contractual licensing and development obligations. This is up from $118.2 mln as of September 30, 2004. In addition to third-party commitments, this obligation includes future payments to sports leagues, movie studios, and other IP holders. Wireless revenue, on the other hand, could ease this margin pressure, although we do not expect its revenue contribution to be meaningful (7% of total revenue) until CY10. JAMDAT’s gross margins trailed downward from an upper 70% range to a high 60% range in CY05. We also note EA – unlike its peers – expenses internal development costs through operating expenses, not cost of goods sold. Consequently, EA’s gross margins are higher than those of Activision, Take-Two, and THQ.
Our operating expense estimates are a function of advertising spend and total headcount. We believe headcount growth accelerates through the cycle, analogous to the increased hiring in CY04 and CY05. We project over 7,700 people (average headcount) by CY10, mostly fueled by additional R&D personnel. Growing development teams and longer development cycles could prevent meaningful and sustained operating margin expansion. Advertising has consistently been 5% to 6% of revenue in the past several years. We are projecting growth in dollar terms but slightly lower in percentage terms. Despite its exclusive licenses, EA still must market its products. We expect operating expenses as a percentage of sales to trend downward as console and portable installed bases grow.
We estimate peak earnings of $2.21 in CY09. Larger console and portable installed bases may fuel this growth as mainstream franchises such as Madden NFL, Need for Speed, FIFA
, and NCAA Football
hit their stride. This implies a peak pro forma operating margin of 20%, below the 27% achieved in the current-gen cycle in CY03.
We project cumulative cycle-over-cycle revenue growth of over 55% for Activision, driven by strength in consoles and portables. Console revenue growth of 59% should be fueled by movie-based IP properties such as Spider-Man
. In the current-gen cycle, Activision had the benefit of two Spider-Man
, one Shrek
, and two other DreamWorks movies. In the next-gen cycle, we expect two Spider-Man
, two Shrek
, and six other DreamWorks movies. Portable revenue growth of 102% should be driven by Activision’s library of mature content that complements the expanding PSP installed base. Portable revenue also faces an easy comp as management released few titles in CY02 and CY03. Titles from id Software should drive PC revenue growth of 33%. We believe Activision’s revenue will fluctuate with the movie releases of Spider-Man
. As such, there could be two revenue peaks, $1.68 bln in 2007 (Spider-Man 3
and Shrek 3
) and $2.22 bln in 2010 (Spider-Man 4 and Shrek 4). We have modeled distribution revenue as a percentage of total revenue; in years with signifi cant releases, distribution as a percentage of revenue should fall.
Gross margins in the current-gen cycle steadily rose and peaked at 41% in 2004. We do not expect Activision to reach peak gross margins of that level in the next-gen cycle. Gross margins will be favorably impacted by the mix shift to higher-margin revenue, such as wireless and micro-transactions. We expect portables to be 10% of total revenue by 2010, down from 12% in 2005. However, offsetting the lower product costs will be the increase in development costs for next-gen titles. IP licensing costs should also trend higher, as we believe it will be increasingly more expensive to license popular content. Therefore, we expect gross margins to peak in 2010 at 39%.
We expect average headcount to grow modestly through the transition period (2006) and accelerate through the next-gen cycle. (Activision confirmed in February 2006 it plans to lay off up to 7% of its workforce.) We project average headcount of approximately 2,600 by 2010, up from 2,000 in 2005. We expect operating expenses per head to fall significantly in 2006, as Activision cuts costs to ride out the transition year. Expenses per head should stay relatively constant, as inflation is offset by outsourcing. Advertising has historically been 7% to 13% of revenue. After bringing advertising costs to new highs in 2005 (13% of revenue), Activision noted during its 3QFY06 earnings call that it expects to better align marketing spend with market conditions. We expect Activision to revert advertising costs to 8% of total revenue through most of the next-gen cycle. Based on our assumptions for headcount and advertising, we expect operating expenses as a percentage of total revenue to trend up through the nextgen cycle.
We believe earnings will peak at $0.81 in 2010. Peak operating margins and peak earnings should result from strong revenue growth, driven by Spider-Man
movie releases in 2010. In the last-gen cycle, peak earnings and operating margins were $0.52 and 14% in 2004.
We project cumulative cycle-over-cycle revenue growth of over 104% for THQ, driven by strength across all segments. Our console revenue growth of 94% is signifi cantly higher than the baseline growth of 32% outlined earlier. The high growth is possible given that THQ’s exclusive rights to Pixar-based games did not start until the fourth year of the currentgen cycle (Finding Nemo
in 2003). Pixar-based games represented 21% of total revenue in FY05 (ends March 2005). We expect four Pixar titles (including Cars
in June 2006) to drive revenue across all platforms from now until the end of the next-gen cycle. Portable growth of 102% is lower than the baseline growth of 143%. One of the major drivers of the baseline growth is the expanding PSP installed base. While THQ may have some successful titles on PSP (ex. WWE
), we do not believe THQ’s library of mostly kids-based games would fi nd a large audience on the platform. Instead, we think THQ’s portable revenue growth will likely come from Nintendo handhelds. We believe revenue should peak in 2010 at $1.52 bln, when the next-gen console installed base peaks and value buyers find THQ titles attractive.
Gross margins in the last cycle have varied from 36% in 2002 to 44% in 2005. We believe gross margins could stay fl at through the next-gen cycle at 42-43%. The continued mix shift to higher-margin revenue should lower product costs. By 2010, we expect portables to be 27% of total revenue, equal to 27% in 2005. However, we believe THQ will not escape rising industry costs that may also plague the other three publishers. Rising software development costs for next-gen titles and higher IP licensing costs may offset any benefit from lower product costs.
We believe operating expenses as a percentage of revenue should trend down through the next-gen cycle, from 36% in 2006 to 31% in 2010. We expect faster revenue growth to offset rising expenses. Our expenses are based on modest headcount growth that accelerates through the next-gen cycle. We project average headcount to grow from approximately 1,500 in 2005 to 2,100 in 2010. We project operating expenses per head to stay relatively constant, as inflation is offset by outsourcing. Our expenses are also based on advertising costs of approximately 7% of total revenue.
Earnings for THQ should improve through the next-gen cycle. Steady revenue growth, driven by more Pixar-based games, flat gross margins, and lower operating expenses as a percentage of total revenue should lead to operating margin expansion. Therefore, we project peak earnings for THQ of ~ $1.80 in 2009-10, with peak revenue occurring in CY10. The peak operating margin of 12% in 2009-10 compares to peak operating margin of 14% in 2001, when peak earnings were $0.67.
We project cumulative cycle-over-cycle revenue growth of 66% for Take-Two, driven by strength in consoles, portables, and PC. We anticipate next-gen Grand Theft Auto (GTA)
and Midnight Club
releases in CY07 and CY09, respectively, which should spark the highest annual growth rates in the next-gen cycle. Both franchises should benefit from a potential second year of PS3 availability in CY07 and a larger expected console installed base by CY09. We believe GTA
can help Take-Two achieve console-over-console growth of 47%, in excess of the 32% baseline estimate. 2K Sports titles could offer a consistent annual revenue foundation, but we are not sure if the titles can successfully compete against EA. With possibly one fewer GTA
console title compared to CY00-CY05, we believe Take-Two will aggressively pursue GTA
titles on the PSP – both original content and ports. PC cycle-over-cycle growth has the potential to improve due to new studios such as Firaxis (Civilization
) and Irrational
Games. We expect peak revenue of $1.98 bln in CY09.
In the current-gen cycle, gross margins peaked at 39% in CY02 when GTA III
and Vice City
were selling well. In the next-gen cycle, we expect gross margin to be relatively similar for Take-Two. Downward pressures are next-gen console royalties and distribution. We believe Take-Two may compensate Rockstar handsomely and these future internal royalties could depress gross margins as GTA
unit sales build through the nextgen cycle. In addition, Take-Two pays licensing royalties to major sports leagues, including the NBA, NHL, and MLB (third-party exclusive contract for Take-Two). We believe these are lower margin franchises relative to GTA, Midnight Club
, and Civilization
. Distribution is normally a low margin business greatly affected by the mix of hardware and software sales. Offsetting these downward margin forces are higher sports pricing, greater studio ownership, and greater mix of higher-margin revenue, such as ingame advertising and micro-transactions.
Take-Two was fairly acquisitive in the current-gen cycle. We cannot predict purchases, but we believe it is likely that headcount growth will accelerate near CY09 and CY10. We project nearly 2,600 people (average headcount) by CY10, mostly from additional R&D personnel. Advertising has consistently been 4% to 6% through most of the currentgen cycle; however, we believe this expense increases as a percentage of sales. The company’s 2K Sports franchises will most likely demand higher advertising dollars, especially for non-exclusive sports such as the NBA and NHL.
We estimate peak earnings of $1.71 in CY09 versus $1.46 in CY02, the peak year in the current-gen cycle. We believe this will be achieved through strong growth in GTA
and Midnight Club
. This implies a pro forma operating margin of 10%, equal to the 10% achieved in the current-gen cycle in CY04, when Grand Theft Auto: San Andreas
shipped. Peak pro forma operating margin of 16% occurred in CY02 thanks to two GTA
[Thanks again to Susquehanna Financial Group analysts Jason Kraft and Chris Kwak for their work reprinted here.]