Dear friends and investors,
For the quarter ending June 30, 2018 Forage Capital (“the Fund”) returned an estimated +2.7%, net of all fees and expenses compared to +3.4% for the S&P 500 Total Return Index. At quarter-end, the portfolio was comprised of 11 companies with a median market cap of ~$27bn and its cash balance was 36% of assets.
I added to our position in Red Hat after the stock corrected by nearly 20% in the days following its earnings release, which revealed lower than expected quarterly revenue guidance on decelerating growth in legacy middleware tools. Looking past a quarter or even a year, the larger point is that enterprises are adopting new ways of marshaling compute resources and developing software in order to compete more effectively in a world where “every company is a technology company”. Applications, once monolithic outputs of waterfall development processes, are being re-configured as collections of of independent functions, continuously and separately developed, tested, and deployed in small batches; hosted across multi-tenant public clouds and on-premise data centers; and accessed by end users across various devices. The resulting fragmentation has brought about flexibility, but also enormous complexity. To manage the ever sprawling topology, IT vendors are introducing new layers of abstraction that, in effect, re-aggregate and organize stuff, and act as critical points of integration.
I believe Red Hat’s OpenShift, a cloud-agnostic platform on which to build and deploy container-based applications, will be one of those critical points, and as OpenShift adoption grows, so too will sales of middleware tools bundled with it. Containers represent an increasingly critical enabling technology and yet underlie a small minority of software development today. OpenShift and other emerging technologies are less than 10% of Red Hat’s total revenue, but are ~doubling y/y even while Red Hat’s established technologies, namely RHEL, the Linux based operating system that comprises the majority of revenue, expand at a low-teens rate. There are many different versions of Linux – the fastest growing operating system today, ubiquitous across data centers – but Red Hat is one of the few enterprise grade distributions, claiming nearly 70% of all installations.
I also added to our position in Ryanair. As I’ve written before, in any given year, you will find at least a half-dozen reasons not to own this stock. In the past, it’s been ATC strikes, terrorism, austerity measures, economic contraction, fuel shocks, low-cost competition from incumbents, low-cost competition from upstarts, foot and mouth disease, the Iraq War, Avian flu, Volcanic ash clouds. Recently, the company has wrestled with scheduling snafus, union recognition, and elevated fuel prices. The additional cost burden from raising pay for existing staff and expanding headcount to meet the company’s long-term passenger growth expectations, will overwhelm incremental revenue growth in the short-term, driving a 10% hit to the company’s after-tax profits in fiscal ‘19.
But, the critical point is that Ryanair’s ability profitably grow traffic over time depends not so much on its absolute costs, but on its costs relative to peers, because the latter is what underpins the recursive process whereby Ryanair profitably undercuts its peers to attract passengers and leverages that traffic to negotiate ever favorable landing fees at airports and volume discounts on aircraft, further reinforcing a cost advantage that can be recycled back into lower passenger fares, and so on. Incumbents with broken balance sheets, complacent cultures, and bloated cost structures cannot replicate this process. Ryanair has never produced a loss in any fiscal year over the last 18 and remains solidly profitable and self-funding. Its balance sheet is clean. Exogenous shocks to the airline industry will have far more deleterious consequences for legacy carriers than for Ryanair and other disciplined low-cost carriers, producing a more accommodative competitive environment for survivors over time, even if near-term profits suffer. Management is expecting €200mn in incremental labor costs next year, translating into an additional €2/passenger (on a cost base of €27/passenger), which still puts Ryanair’s ex. fuel unit cost 25% below that of next lowest cost European carrier, Wizz Air, and at a fraction of those of incumbents like IAG.
The most significant detractor of performance this quarter was MercadoLibre, whose stock declined 16% during the quarter. I did not add to our position, however, as some of the competitive concerns here are valid and my confidence in the long-term viability of this company is properly captured in its position size, which was just under 5% of the portfolio at quarter-end. In response to an increasingly competitive landscape in Brazil – not just the intensified efforts of Amazon but also those of domestic brick and mortar retail incumbents – MercadoLibre has been ramping up shipping subsidies, to the great detriment of near-term profitability. Also, earlier this year, Brazil’s national postal carrier hiked its shipping rates, directly eating into MELI’s margins for most of the quarter.
MercadoLibre is not in denial about where its logistics depth needs to be. The company is now operating its own distribution centers in Mexico and Brazil, and has another one coming online in Argentina. I would expect in-house fulfillment, while a small percentage of Libre’s orders today, to increasingly cannibalize its third party drop ship volume. Furthermore, product breadth is still a key pillar of competitive differentiation – that 1P competitors are so actively pushing marketplace platforms suggests as much – and relative to peers, Libre has a commanding lead on seller engagement.
And then there’s MercadoPago, the company’s online payments solution that intermediates 90% of Libre’s GMV (basically, it is to MercadoLibre what PayPal once was to eBay). Even now, but particularly 15 years ago when MercadoLibre launched Pago, access to financial services was limited and transacting online was still a novel undertaking, fraught with uncertainty and fraud. By stepping in between transactions and taking control of the payments flow (basically, acting as an escrow agent), MELI could ensure that sellers got paid for the items they shipped and buyers received the items they paid for. Pago helped spur Libre buy-in from merchants and consumers, and symbiotically, Libre – a high velocity marketplace with a real need for payments intermediation – served as an ideal first use case for Pago. The interaction between the two is a force multiplier to the cross-side network effects that characterize the business models of both Libre and Pago individually. Pago now has a life of its own, with 2017 marking the first year that total payment volumes on Pago exceeded the GMV on Libre. Along with market leader PagSeguro, Pago is favorably positioned to gain share from incumbent card processors and fintech start-ups, and given the vast addressable market, there is a decent chance that Pago is eventually worth more than Libre.
Selection, logistics, payments. Those are the core pillars required to support a viable marketplace in an emerging economy. Providing credit also helps, and in that vein MercadoLibre recently rolled out MercadoCredito, offering working capital loans to merchants on the Libre marketplace, first in Argentina (4q16), then in Brazil and Mexico (2017). With only ~$90mn in merchant receivables, Credito is still a tiny business, but one that deserves more emphasis for the following reasons: 1/ small merchants in Brazil are often capital constrained and locked out of traditional banking channels. Credit keeps merchants on the Libre platform and gives them the means to supply more inventory; 2/ because MELI has a real time view into merchant sales activity and intermediates transactions, it is far better positioned than outside lenders to intelligently underwrite merchant loans (Square and PayPal tout similar advantages) and collect payments; and 3/ the returns are compelling, with ~40% APRs covering low-single digit loss rates many times over.
Please check out the scuttleblurb blog for more detail:
 The price points of OpenShift and OpenStack (Red Hat’s private cloud solution) are many times greater than standalone RHEL (Red Hat’s flagship OS), and as the percentage of Red Hat’s deals embedded with these technologies has dramatically expanded – from 5% a year ago quarter to 19% last quarter – so too have the company’s average the deal sizes, with number of $1mn+ engagements growing nearly 50% over last year.