June 2019 Investment Update

Updated: Jul 31, 2019

Dear investors,

We had a disappointing month, contracting -7.4%, which reduced our year-to-date return to 12%. Shorts and hedges, which contributed 9.5% in May, contracted -8.8%, while familiar names of Carvana (+1.1%), Lundin Petroleum (+0.7%), Stanmore Coal (+0.5%), iQiyi (+0.4%) and Afterpay (+0.4%) made the largest positive contributions.

Position update

Carvana is the largest position after a fairly significant price appreciation. The stock jumped from our entry level of $35 to $76 in May, before dropping down to ~$59 as I write. Since initiation, Carvana announced growth of 110%, well ahead of expectations for 2019 of 70%. Pleasingly, margins continued to rise, demonstrating the operational leverage in the business. The firm issued its first securitisation of self-originated loans, and increased the number of cars traded in by customers, which are refurbished or sold on to wholesale markets for incremental margin. This gives a road map for gross margin per car to rise above management’s mid-term target of $3000.

The proposition for consumers remains compelling: a fixed, no haggle price, a 7 day test-drive/return period, a 100-day warranty, and no reason to deal with car salesmen. Long term, the firm should have a significant cost advantage over existing dealerships, as there’s no need to pay sales incentives or invest in retail real estate. We also view the ability to hold a comprehensive nation-wide inventory, rather than the few cars that can fit in a dealership, as an intrinsically better proposition for consumers than traditional local dealers.

At target profitability and 2% penetration in the United States - which Carvana has achieved in some markets - the company would be worth well in excess of 100% of its current valuation. The second hand car market is highly fragmented, so we think it unlikely that Carvana would stop there.

We have covered Pinduoduo in a previous note. At current rates of growth, Pinduoduo will end up with over 20% market share of online retail in China, (the firm has grown from zero to 8% in the past few years) and be worth 5x from current levels. If the firm can reach comparable levels of profitability to Alibaba, this could be 10x from here. This would add 30-60% to the fund, and handily justify the current volatility.

Our only other Chinese (or indeed Asian) investment is iQiyi, which recently reached the milestone of over 100 million subscribers, well ahead of 67 million subscribers when we first invested. Unlike Netflix, iQiyi has a freemium model, and earns advertising revenues from most of its ~500 million monthly active viewers. Our investment thesis was predicated on the assumption that these free viewers would steadily convert into subscribers, and this has so far proven the case.

In their latest results, the firm announced annual membership growth of 64%, content distribution growth at 66%, and other revenue such as games, increasing at >143%. This ‘other revenue’ now accounts for 14% of the total, while advertising, which was roughly equal to membership revenue at purchase, has become a smaller and smaller piece of the pie. Advertising revenue was flat year-on-year, and this has been part of the ‘disappointment’.

As fast-growing subscription, live content, talent management and gaming revenues become more and more important, growth may even accelerate from here. These alternative revenue streams are largely incremental to content spend, and should allow iQiyi to reduce content costs as a percentage of total revenue. We think this the key metric to watch going forward, and are happy to play a long game.

While iQiyi’s subscriber and revenue performance is well ahead of our initial modelling, the revenue multiple has contracted substantially, and is well below its peak earlier this year, at about 3.5x trailing sales. This is a fraction of where subscription technology and media businesses are currently trading elsewhere (Netflix is at 8.2x, with substantially lower growth).

Multiples wax and wane, and when they next increase, the company will price on a substantially higher base of subscribers and revenues.

Both of our remaining Chinese investments have the potential to increase several-fold over the coming years – and are more than 50% away from where they were trading only a few months ago, before the reignition of trade disputes. Both companies are entirely focused on the Chinese domestic consumer, so while sentiment certainly drives the day-to-day and month-to-month price action, fundamentals will be largely unaffected by trade disputes.

Afterpay added a significant bite to performance in the final hours of the month. The stock is up more than four-fold from our initial purchase. The significant developments were the imposition of an external audit by Austrac, and the release of a free version of buy-now-pay-later by Klarna. We view the announcement regarding Visa's development as a pay-later API as something of a red herring, as they are simply creating the tools to allow banks to offer pay-later models, which they have always been able to do. That competitive threat was always there, but fortunately payment technologies constitute one of the largest markets with plenty of room for multiple competitors and business models. The biggest risk here is that buy-now-pay-later becomes a loss leader for financial firms and used to acquire customers. In this scenario, Afterpay will have to continue to drive incremental retail sales to keep merchants on side.

There has been greater stock dilution than the market seems to have expected, though less than we originally modelled. We take very conservative assumptions around dilution, and will only concern ourselves here if Afterpay exceeds these.

Ignoring the noise, long-term performance will be driven by fundamental KPIs, specifically:

  1. The total merchandise value transacted in each country. This can be derived from reported user numbers, transaction frequencey and average transaction value.

  2. Net transaction margin

  3. Staff costs & other operational expenses.

The firm can be valued conventionally using these metrics, and we view competition as relevant to the extent these fundamentals are affected. There has been encouraging initial take-up in the United Kingdom, and momentum in the United States remains strong.

We have used the rally since the end of June to trim the position somewhat, but this remains a large position.

Bluebird bio has come off its recent lows but remains ~75% away from its peak in 2018. Previous buy-outs of gene therapy companies imply a substantially higher valuation still. Our biotech investments have driven some of our recent underperformance, as many of these peaked in early or mid 2018 and have yet to recover.

The EU launch date of Zynteglo, bluebird’s first treatment, is steadily approaching. Priced at $1.8m per successfully treated patient, Zynteglo will launch in the US in 2020.

Bluebird’s oncology collaboration with Celgene remains best in class, and in it was a shame that Celgene, a natural buyer of the firm, was itself bought out by Bristol Myers-Squibb. Goldman Sachs has an M&A price target about 3x from where the stock is now, and irrespective of corporate activity, we believe this is a realistic, if optimistic, mid-term target, as bluebird transitions from a clinical research organisation to a revenue producing best-in-class gene therapy platform.

Bluebird bio has $1.9 billion of cash, which should see it through to 2022, by which time we expect the firm to have multiple approved treatments.

One of our more interesting biotech investments is Oxford Biomedica, a firm with leading expertise in lentiviruses, which also form the core of bluebird bio’s technology. Gene therapy generally utilises two groups of viruses: lentiviruses, a group that includes HIV; and adenoviruses, a class that includes the common cold. Lentiviruses have significant advantages, such as being able to carry higher genetic payloads and infiltrate non-replicating cells.

Oxford Biomedica’s business model is quite different from typical drug development companies. Instead of funding clinical research on favoured targets, they develop early-stage gene therapies, then license those out to other firms, while retaining production facilities in-house.

The other firms then take on the development costs and risks, and pay Oxford Biomedica for the production of viral factors. Oxford Biomedica then earns revenues throughout clinical development, and on approval, earns further royalties and production revenues. Oxford Biomedica currently has the only globally approved facility for manufacturing lentiviruses.

One of the best measures of quality of a biotech company is the calibre of partners it can attract, and Oxford Biomedica’s partners are truly top class. Novartis’s Kymriah treatment is actually one of Oxford Biomedica’s and how we originally became aware of the company:

Their product pipeline is significant, with a number of new gene therapies ready for partnership.

We believe this field will expand significantly, and we’re not alone: the FDA expects to approve 10-20 gene therapies a year by 2023. Oxford Biomedica has a capital-lite model, but has significant upside in a large and growing suite of products. Oxford Biomedica is about 9% below our entry price, but about 50% from its mid 2018 high.

Unlike most biotech companies with drug development upside, Oxford Biomedica is already profitable, with revenues rising 72% in 2018, and EBITDA rising from minus £1.9m to +£13.4m. The firm has net cash of £32.2, and a market cap of £522m

We re-entered Google after the stock fell significantly, around the time it was reported the US Department of Justice was preparing an antitrust investigation. Should Google be forced to spin out YouTube, Google Maps or some of its other core properties, we believe this will create significant shareholder value. Conglomerates generally only benefit the very top level of management. Consumers, shareholders and mid-level executives would all likely reap rewards from a break-up. 2018 revenue growth slowed to 23%, and is expected to slow further still by the end of 2019, but valuation contracted to below 10x forward EV/EBITDA.

Given the quality of the business, potential break-up value, and net balance sheet cash, we think this is an excellent time to re-enter one of the world’s top companies.

Amazon is one of our largest positions, and after a strong rally, is now about 8% of the portfolio. Amazon’s AWS service may well be one of the best subscription businesses around, with high switching costs and extraordinary flexibility and value for customers. Amazon could significantly increase advertising revenues and might already be considered Google’s top competitor in search. Google also has significant underutilised value, and has barely monetised a number of applications with over a billion users. For example, Google Maps forms the backbone of many of the world’s most popular apps and was given away almost for free until 2018. As an aside, we, like many others, have found ourselves paying monthly subscription revenues to Microsoft, Google, Amazon, Netflix and Apple. Quite an interesting shift from only a few years ago, and a key driver of value creation in global markets. One of our more significant errors has been to be underexposed to this shift.

We entered Autodesk, a firm that makes design, engineering and construction software used by builders around the world. There software has significant lock-in effects, as expertise develops around a specific tool stack. We paid up here, at 9.6x forward sales, but the firm is already profitable, and trades at a forward EV/EBITDA of 28x. This EBITDA was up 138% over the past 12 months. We have a buy list of about 20 stocks, of which companies like Autodesk make up a significant proportion.


A quite possible outcome is continued weakness and contraction in industrial and cyclical sectors, but no outright collapse. This would prompt lower rates, but quite possibly higher valuations for capital-lite, rapidly growing companies, for which macro considerations like GDP are largely irrelevant.

There is fairly considerably upside in the portfolio right now. We expect Pinduoduo to continue taking 25% of the e-commerce market, and in the upside case, be worth >$200b over time, which would be a 10x return from here. This alone could add over 60% to our portfolio, and dwarf the ~-1% impact from the fall since our initial purchase. It’s increasingly likely that bluebird bio executes on its four core programs, and if so could add more than 8% to the portfolio, and more again if targeted in M&A. Three of our previous gene therapy players were bought out at higher valuations with weaker pipelines and lower platform value. If Afterpay continues accreting customers at its current rate, we expect the firm to be a US$10b company within two years – which would add an additional 15% to the portfolio, This would put to shade even the sharp volatility of the end of the month. At some point, we expect Stanmore coal to trade at 5x EBITDA, which would add another 8-10% to the portfolio. If this proves optimistic, we expect to generate our return though sheer cash generation - over the past three months alone Stanmore announced operations had increased cash from $58m to $90m, equivalent to about 8% of their current market capitalization. And if Carvana, our largest position, reaches 2% market penetration in the US, this could add 20-25% to the portfolio, and multiples of this over the long term.

As always, we appreciate your patience and support, and hope the above offers some comfort in the quality of the opportunities, and the rewards on offer for long-term investors.

Best wishes Michael