CoinShares — Maintaining robust earnings in Q321

We consider CS’s total comprehensive income and adjusted EBITDA as key financial performance indicators that are more relevant than its net income (see our initiation note for details). CS reported a solid total comprehensive income of £26.2m in Q321, significantly up from £2.7m in Q320 and close to £26.6m reported in Q221. Similarly, the company’s adjusted EBITDA reached £26.0m compared to £4.7m in Q320 and £28.6m in Q221.

CS’s management fees were £18.4m in Q321, down c 6% versus £19.6m in Q221 (but still up 3x vs Q320). Consequently, Q321 fees represent the second-best result in CS’s history after Q221. The sequential decline was a function of lower average digital asset prices, bitcoin (BTC) and ether in particular, as a result of the sell-off in May 2021 (despite the subsequent partial rebound that started in late July 2021). Moreover, CS’s XBT Provider Trackers saw net outflows of c US$330–340m in Q321 (according to our estimates, based on company data). This was partially offset by net inflows into the CoinShares Physical ETPs, which are still in a ramp-up phase and thus remain only a minor top-line contributor for now (£0.2m in Q321). After the acquisition of the ETF index business from Elwood Technologies in July 2021, CS started recognising management fees from the Invesco CoinShares Blockchain Global Equity UCITS ETF, which in Q321 amounted to c £0.6m. We note the index business has recently expanded through feeder funds in Japan and Thailand (with the launch of another feeder fund in India scheduled for November).

CS’s CSCM operations posted income/gains of £8.4m in Q321, visibly up from £3.2m in Q320, but down from £14.7m in Q221. We note, however, that Q221 (and Q121) was characterised by particularly high market activity and volatility (which normally benefits CSCM), while Q321 has been a more muted period (eg BTC trading volumes went down by c 41% according to Arcane Crypto), which may be at least partially due to seasonality factors (summer holiday season). Consequently, CSCM’s delta neutral trading strategies delivered a £2.6m gain (vs £1.6m in Q320 and £9.1m in Q221). Meanwhile, fixed income activities generated a £3.1m income in Q321 (up from £1.0m in Q320 and £1.7m in Q221). CS’s liquidity provisioning income (generated primarily on XBT Provider Trackers) stood at £1.7m vs £0.7m in Q320 and £3.3m in Q221. CS continues to explore new trading opportunities and management highlighted during the Q321 investor call that CSCM started to engage in yield farming (ie staking or locking up digital assets to generate tokenised rewards) on decentralised finance protocols such as Compound, Aave and Maple Finance. According to a Glassnode analysis from early September 2021, USDC lending rates on Compound and Aave stood at c 3–4% in late August and early September 2021, but were as high as 12% before the market sell-off in May 2021. CS has also started to participate in the governance of selected protocols (eg Aave).

CS also recognised c £4.8m gains on its principal investments, most notably from 3iQ (£2.0m), Solana tokens (£1.6m) and carried interest from CoinShares Fund II (£1.4m).The latter was mostly driven by the US$155m series B funding round of the blockchain infrastructure-as-a-service company Blockdaemon (which is among others a node operator and staking infrastructure provider), carried out at a valuation of US$1,255m. Management highlighted that the current carrying value of its stake in 3iQ is still below the valuation implied by its latest funding round. After balance sheet date, the company announced a US$11.8m investment in a 9.02% stake in FlowBank, a Swiss-based online bank (see our initiation note for details).

CS’s administrative expenses fell from £9.8m in Q221 to £6.2m in Q321 (thus returning closer to the Q121 level of £5.8m), which the management considers a more appropriate run-rate going forward. CS continues to improve its tech infrastructure and has added a number of new engineering hires to the team (with the total number of employees at CS of 58 compared to 42 at end-2020).

CS’s cash position at end-September 2021 was £18.8m, down from £53.3m at end-June 2021. Based on our discussion with the management, we understand that a significant part of the previous cash position has been deployed into CSCM’s trading strategies, for example a cash and carry strategy as the digital asset futures markets moved from backwardation to contango in Q321. The latter was also an important driver of the net amounts due to brokers, which was £122.2m at end-September 2021 (including utilised credit lines CS has with its brokers) versus a net amount from brokers of £24.9m at end-June 2021. However, this movement was offset by increased digital asset holdings purchased to hedge these liabilities.

XBT Provider is operating under an exemption granted by Nasdaq (which operates the exchange where the XBT Provider Trackers are listed) given it does not hold an EEA MiFiD-compliant regulated status. To take advantage of the exemption, it has been required to provide a guarantee for its activities (which so far was provided by a group subsidiary, CoinShares Jersey). CS’s intention was to acquire Peak AM Securities (which holds an EEA MiFiD-compliant regulated status), but acquisition negotiations were terminated in August 2021 (as the Swedish regulator denied CS’s application for the ownership suitability assessment in relation to the acquisition of Peak AM Securities). However, CS’s legal and compliance team identified an alternative solution to transfer the guarantee internally to CoinShares Capital Markets Jersey. As this entity is responsible for hedging the liability under the programme, CS’s clients are now offered a more transparent setup with the hedging and guarantee related to the XBT Provider products within a single entity. This secures the continuity of the product platform while satisfying the requirements of the Jersey regulator.

The XBT Provider Trackers saw c US$1.1bn of net outflows ytd to 29 October 2021 (which resulted in the release of c £11m accrued management fees, according to the management), which we believe has been primarily driven by earlier investors realising gains following the strong bull run in digital assets since March 2020. We note, however, that net outflows eased somewhat recently with c US$171m between August and October. There were even a few weeks with net inflows recorded in late September and early October, while in November so far, the products reported a US$3.3m net inflow, according to CS’s weekly fund flow reports. Management also highlighted the number of individual holders of XBT Provider Trackers tripled from c 24k at the beginning of the year to 72k currently, suggesting these products continue to attract a considerable number of new investors.

Meanwhile, net inflows to the CoinShares Physical ETPs continued with c US$70m in October (c US$180m ytd after excluding CS’s seed assets), offsetting net outflows from XBT Provider Trackers during the month. According to the management, CoinShares Physical was the third-best ETP platform in Europe in terms of inflows in Q321 (up from fifth in Q121 and Q221) out of nine digital asset ETP issuers active in Europe. Inflows should be supported by listing on further exchanges (eg the Bitcoin and Ethereum ETPs were recently listed on Euronext Paris and Amsterdam). CoinShares Physical ETPs have been introduced to appeal to institutional investors and we believe growing institutional adoption should represent a positive driver for fund inflows. This seems to be confirmed by the Institutional Investor Digital Assets Study conducted by Fidelity Digital Assets between 2 December 2020 and 2 April 2021 (published in September 2021) on a group of financial advisers, high-net-worth Individuals, family offices, pension funds and defined benefit plans, crypto hedge funds and venture capital funds, and more traditional hedge funds and endowments and foundations. According to the study, the share of surveyed European investors who intend to purchase digital assets in the future increased from 60% a year earlier to 75% currently. Interestingly, the study showed a more progressive view towards digital assets of European investors compared to US investors. Also important is that 84% of US and European investors surveyed indicated they would be interested in institutional investment products that hold digital assets. Their allocations to such investment products increased from 14% to 29%, although the share of surveyed European investors holding digital assets directly has also increased versus the prior year, from 29% to 41%.

On the back of the more limited net outflows from XBT Provider Trackers recently, we have reduced our assumed net outflows in FY21e and FY22e to c US$1,178m and US$1,023m, respectively (vs previous assumptions at US$1,319m and US$1,302m, respectively). We have also slightly raised our global digital asset allocation assumptions for FY21e to FY24e, supported by the recent solid inflows across the market, and recent strong performance of digital assets. Nevertheless, we keep our relatively conservative 2.0% and 2.5% allocation forecast for FY25e and FY30e unchanged. The above has translated into higher management fee forecasts for CS of £100.1m in FY22e (vs £72.0m previously) and £92.9m in FY23e (vs £66.5m previously), while our FY21e forecast remains broadly unchanged (£77.6m vs £78.1m previously). That said, as fees generated on XBT Provider Trackers are accrued until investors redeem the certificates, the above had only a limited positive effect on our forecasted cash release of fees in the near term (although it assisted the expected cash fees in later years). Moreover, the reduced forecast outflows (together with an adjustment we made to the assumed average spread) translated into a lower expected income from liquidity provisioning in the near term.

Given the large swings between quarters in CS’s amounts due from and to brokers, including the utilised proportion of credit lines with brokers, we are now reflecting the interest expense on these credit lines in our free cash flow calculations rather than adjusting CS’s valuation for the last reported net amounts due from/to brokers. Moreover, we have reflected the lower cash position at end-September 2021 compared to end-June 2021 and updated our fx assumptions. As a result, our value estimate of CS in our base-case scenario is SEK119.8 per share (vs SEK120.3 previously). Since our initiation of coverage on 15 October 2021, CS’s share price has appreciated c 16% to SEK87.2, which still leaves an upside potential of c 37% based on our fair value estimate. We have also updated our ‘Crypto Winter 2022’ scenario (see our initiation note for details), which now implies a fair value per share of SEK72.9.