CoinShares International — Q1 results affected by subdued crypto sentiment

Before we examine the Q122 results, we discuss the impact of the recent collapse of the TerraUSD (UST) token on CS’s Q222 results. UST was a stablecoin (ie a cryptocurrency with its value pegged to another asset, in this case the US dollar). Contrary to centralised stablecoins such as Tether (USDT) and USD Coin (USDC), where the peg is facilitated by off-chain reserves held in US$-denominated traditional assets (eg cash and US Treasuries), UST’s peg was maintained through an algorithm executed by a smart contract (ie a blockchain-based computer programme) and initially underpinned exclusively by the value of LUNA, the native token of the Terra blockchain, which hosts the UST. Each UST token could be issued to/redeemed by investors in exchange for one US dollar worth of burned/issued LUNA tokens. Consequently, any deviation from the peg has (until recently) been arbitraged out. While the Luna Foundation Guard (a non-profit entity developing and governing the Terra network, including UST) created reserves in bitcoin (BTC) and selected other digital assets (eg AVAX) earlier this year to defend the peg, UST was still to a large extent reliant on the value of LUNA.

Demand for UST came primarily from users of the decentralised lending protocol Anchor (hosted on the same blockchain, ie Terra), which CS also participated in as a lender, which it disclosed for the first time in its Q122 results published on 3 May 2022. Anchor offered a very attractive fixed deposit rate of c 19.5% pa, which at its peak on 5 May 2022 attracted more than US$17bn (according to defillama.com), significantly in excess of the amounts borrowed from the protocol (and at rates considerably below the 19.5%). This created an interest income shortfall, which had been covered by a so-called yield reserve (which had to be regularly topped up), hence the deposit rate was unsustainable in the long run. The value of native tokens of blockchains hosting smart contracts is normally underpinned by the demand for computing power on the network and the amount of funds deposited in its DeFi applications (referred to as total value locked, or TVL). Given that Anchor represented more than 50% of Terra’s TVL before the UST collapse, there was a significant degree of ‘circular backing’ in the system, making it vulnerable to a ‘death spiral’ scenario in which massive withdrawals from the Anchor Protocol, and the resulting supply of UST tokens, would lead to large, dilutive issuance of new LUNA tokens triggering a fall in LUNA’s price, leading to more withdrawals from Anchor and UST selling pressure, eventually ending with UST de-pegging from the US dollar. This scenario unfolded after 6 May 2022, with LUNA’s price falling from c US$100 to close to zero and UST’s price now at a mere US$0.09.

In the Q122 report, CS highlighted that it deployed US$313m across a number of DeFi protocols, which met the company’s risk-adjusted reward expectations, including Anchor, TrueFi and Maple Finance. The UST de-pegging unfolded gradually, with the Luna Foundation Guard initially trying to defend the peg using its reserves (mostly BTC holdings). As a result, UST traded at or above US$0.90 for most of the time during the two initial days of the de-pegging (after a temporary dip to c US$0.70) before the collapse in subsequent days. This gave CS enough time to withdraw its Anchor deposits before the final meltdown. Still, the company highlighted in its 2021 annual report that it incurred an exceptional loss of £17m on liquidating its UST holdings.

We note that the other two major lending platforms mentioned by CS in its Q122 report are not based on algorithmic stablecoins, but on the centralised stablecoin USDC, the reserves of which are exclusively held in cash and equivalents. TrueFi is a platform offering uncollateralised lending based on on-chain credit scores with an annual percentage yield (APY) of c 9–13% on most of its lending marketplaces and pools (from a combination of interest and native token rewards). Maple Finance is an institutional-grade protocol for under-collateralised lending with an APY of c 7–11% on most of its lending pools.

CS reported adjusted EBITDA of £18.7m in Q122, visibly down from the exceptionally strong £34.2m in Q121 (and £32.9m in Q421), mostly as a result of lower capital markets infrastructure (CSCM) income/gains and higher operating expenses (see details below). Consequently, its adjusted net income came in at £16.1m versus £32.5m in Q121 and £29.7m in Q421.

CSCM income/gains stood at £10.2m compared with £22.8m in Q121 and £16.2m in Q421. Most notably, CS’s gains on delta neutral trading strategies were quite weak at £0.6m (vs £10.1m during the Q121 ‘crypto hype’ and £5.3m in Q421) and fixed income activities generated income of only £1.2m (vs £2.7m in Q121 and £3.5m in Q421). This was due to lower digital asset market volatility (which is unfavourable for CS’s delta neutral strategies), lower trading volumes (BTC turnover down 60% in Q122 vs Q121, according to CoinText) and a flat term structure (limiting eg ‘cash and carry’ opportunities). More muted market activity has also affected liquidity provisioning income, which stood at £2.0m in Q122 versus £6.3m in Q121 and £2.4m in Q421. Amid more limited opportunities for the above strategies, CSCM’s attention shifted to the DeFi ecosystem, where it generated solid income and gains of £6.3m in Q122 versus £3.6m in Q421 (CS entered the DeFi space in Q321).

CS’s asset management fees reached £17.2m in Q122, broadly comparable to the £17.1m reported in Q121, but down from £25.5m in Q421 as a result of the decline in crypto asset prices, which started in November 2021. The company is gradually shifting its fee revenue away from the legacy XBT Provider products, although these still generated £16.3m in fees in Q122, with CoinShares Physical products and the Block Index contributing a further £0.4m and £0.5m, respectively. Overall, CS AUM stood at £3.95bn at end-March 2022 versus £3.36bn at end-March 2021 and £4.18bn at end-2021. Here, we note that the Q121 figure does not include Block Index AUM, as the business was acquired in Q321 (CS’s ETP AUM was £3.07bn at end-March 2022). CS is seeking to further expand the geographical footprint of its ETPs, currently focusing on ‘last mile’ platforms such as Finanzen.net zero it partnered with last year.

CS booked a minor £0.15m loss on its principal investments (vs a minor £0.02m gain in Q121), with some of the holdings affected by the decline in digital asset prices during the quarter. The £2.36m carried interest on CoinShares Fund II (earned primarily on the back of Blockdaemon’s successful Series C funding round) was more than offset by, among other things, the downward revaluation of 3iQ (£1.6m), Solana tokens (£0.6m) and DIKO tokens (£0.2m). In Q122, CS increased its stake in FlowBank from 9.02% to 29.3% (which represents voting rights equal to 32.06%) by investing CHF24.7m – see our initiation note for details on FlowBank. Moreover, it made some other minor investments in Impervious (£380.6k), Pocket Network (£223.5k) and Alliance Labs (£184.1k).

CS’s direct costs (consisting primarily of custody, trading and issuer fees) went up to £3.3m in Q122 from £1.9m in Q121. This includes £1.6m in the asset management segment, up versus £1.3m in Q121 amid new ETP launches, which reduced the gross profit margin of the asset management platform division from 93% in Q121 to 91% in Q122. The capital markets segment posted direct costs (excluding interest expenses) of £1.4m in Q122 versus £0.6m in Q121, affected by an increase in average trade execution costs. This was coupled with higher interest expense of £2.2m in Q122 versus £0.7m in Q121, arising from higher utilisation of CS’s credit lines with brokers.

Other administrative expenses went up to £6.5m in Q122 versus £3.9m in Q121, which includes £1.2m spent in the consumer platform segment, that is Napoleon Crypto, acquired in Q421. CS highlighted that it is making good progress in the integration of the business and hopes this will be reflected in the results throughout the rest of the year. CS is developing Napoleon’s automated trading bots (NapBots) to support business expansion beyond traders to capture the interest of long-term investors seeking access to digital asset strategies. For now, the consumer platform reported a £0.7m operating loss in Q122. Operating expenses also rose in the asset management and capital markets segments as CS scales up its business. As a result, total headcount across the CS group rose to 95 at end-March 2022 versus 44 at end-March 2021. As an example, CS established a marketing and design team led by Benoit Pellevoizin (previously VP of marketing at Ledger, one of the top crypto security and infrastructure providers) who joined CS as head of marketing and communications.

As discussed in our Q421 update note, CS received regulatory approval to add a further 46 digital assets to the main CoinShares Physical prospectus (on top of the four ETPs launched in 2021) and began launching new physically backed ETPs in 2022. It has started with native tokens of top proof-of-stake blockchains, such as Tezos, Polkadot and Cardano (described in our Q421 note), as well as Solana. The latter is particularly interesting as it was launched in partnership with FTX, one of the top global crypto exchanges, representing another opportunity for both players to tap into the institutional investors space. Similar to the previous physical staked ETPs, CS’s Solana ETP offers a combination of reduced management fee (to 0.0%) and stable staking rewards at 3.0% pa. The current staking yield on the Solana blockchain stands at c 5% pa (according to stakingrewards.com) and CS will retain rewards in excess of the above-mentioned 3.0% pa. The current AUM of the Solana ETP stands at c US$46m (as at 31 May 2022), of which the vast majority are seed assets provided by FTX (1m SOL tokens). Moreover, CS launched Chainlink, UniSwap and FTX Token ETPs (the latter having c US$40m of seed assets provided by FTX).

We estimate that the total AUM of Western European digital asset ETPs stood at US$9.2bn at end-March 2022, somewhat down from US$9.7bn at end-2021, due to a combination of lower digital asset prices and net outflows (but still up significantly from US$3.1bn at end-2020). Outflows from XBT Provider trackers continued ytd to 27 May 2022 at c US$344m (likely driven by further profit taking), according to CS’s weekly digital asset fund flows reports. Meanwhile, CoinShares Physical ETPs saw notable net inflows ytd of c US$86m (excluding seed assets), even despite the crypto bear market, which suggests that this product group is gaining traction. This compares with net inflows of c US$57m for 21Shares and net outflows of US$147m for ETC Group, CS’s two major European competitors by AUM. We estimate that net inflows into CS’s physically backed ETPs were still primarily driven by its largest products based on bitcoin (BTC) and Ether (ETH).

Finally, the Invesco CoinShares Global Blockchain UCITS ETF saw net inflows of US$47m to 27 May 2022 (according to CS’s above-mentioned weekly fund flow reports), which represents c 29% of total net inflows into blockchain equity ETPs so far this year. This brought its AUM to US$607m as at 27 May 2022, which means it is now the second largest blockchain equity ETF and, together with the first player, it covers c 67% of the total segment’s AUM globally.

The sentiment towards digital assets remains muted amid elevated geopolitical and macroeconomic risks, monetary tightening across a number of regions (most notably in the United States), as well as the UST collapse. Consequently, the price of major digital assets such as BTC and ETH retraced from end-March 2022 levels (reached following the upward price move in the second half of March). Crypto markets have also exhibited a relatively high positive correlation to tech indices such as the Nasdaq Composite Index, which is down c 23% ytd. As a result, we have decided to adjust our near-term assumptions for global allocation to digital assets to 0.4% and 0.8% in FY22 and FY23 from the previous c 1.21% and 1.43%, respectively. Our new assumptions imply a BTC and ETH price at end-2022 of c US$16,600 and c US$1,200, respectively (vs the previous US$48,300 and US$3,700, respectively). This compares with the current BTC and ETH price of US$31,600 and US$1,940, respectively. We now assume net outflows from XBT Provider products of c US$0.5bn in both FY22 and FY23 (vs net outflows of US$0.9bn and US$0.8bn previously), corresponding to a cash fee release of £40m in each year (compared to £28.8m posted in FY21). Despite the negative impact of lower crypto prices on AUM, we expect the CoinShares Physical platform to post solid net inflows of US$312m in FY22 (including the US$90m and US$40m seed assets of CS’s Solana and FTX Token ETP, respectively) and c US$175m in FY23 (vs previous US$250m and US$450m, respectively). We do not account for any meaningful seed assets of newly launched ETPs in FY23. This translates into our management fee forecast of £51.9m in FY22 (£81.3m previously) and £37.1m in FY23 (£82.5m previously).

Moreover, we have significantly reduced our CSCM income/gains expectations for FY22e to factor in the £17m exceptional loss from the UST stablecoin collapse in Q222. We have also reduced our expectations for FY23e and FY24e to reflect lower income from liquidity provisioning (which is a function of XBT Provider’s AUM and inflows/outflows) and fixed income activities (as we expect CS to focus more on DeFi). We assume much lower gains from delta neutral strategies this year (given the weak Q122 and probably also the beginning of Q222) but expect a rebound in subsequent years. We now recognise the excess staking rewards CS will earn on its staked ETPs in other CSCM gains (previously we included it in revenues). We now forecast adjusted EBITDA of £41.0m in FY22 and £42.6m FY23 (after the exceptionally strong £121.7m in FY21). Our revised forecasts in the base case scenario imply a fair value per share of SEK85.0 (compared to our last valuation of SEK117.2 per share), which currently offers c 113% upside potential. We estimate that the XBT Provider fees accrued so far by CS and its gross cash at end-Q122 cover c 70% of the company’s current market cap.

Despite the significant decline in digital asset prices from the November 2021 peak, we continue valuing CS in our base case scenario with the assumption of resumed secular trend of increasing global allocations to digital assets to 2.0% by 2025 and 2.5% by 2030 (which is the key driver within our valuation model). As discussed in the sensitivities section of our initiation note, digital asset prices tend to be volatile and major bull runs occurred after every Bitcoin halving (in 2012, 2016 and 2020) and lasted for one to 1.5 years in the former two cases. The start of the current crypto bear market (c 1.5 years from the May 2020 halving) is in line with the above pattern. However, we have not seen enough indication that global digital asset adoption has been derailed (eg heavy crypto ETP/ blockchain ETF outflows, suspension of new product launches) to apply our ‘Crypto Winter 2022’ scenario outlined in the initiation note in full. A notable example of further interest in the digital asset space is for instance the recent launch of thematic ETFs by Fidelity and BlackRock. Nevertheless, we will closely monitor the developments throughout the rest of 2022 for signs of a halt in crypto adoption. Moreover, we have reduced our assumption of digital asset ETFs/ETPs as a percentage of total European digital assets allocation from 5.0% to 4.0% by FY25e to reflect more conservative forecasts of fund inflows in the near term.