• The bankruptcy of BlockFi adds another layer to the wreckage deepening the crypto winter, but there are real use cases for the blockchain technology that mean that at some point it is almost certain to rise again and take its proper place in the economy.
• For those with a long-term view and a contrarian nature, I suspect that we are not far from the bottom, but it may be a very long wait for the turn.
• In the world of crypto, it feels like the world is coming to an end (as it did in 2008) but all that is really happening is that valuations are correcting, and businesses built on nothing are failing.
• Bitcoin is a good example of this as at around $16,000 it is not that far above what it costs to create meaning that supply will slow down and provide support to the price.
• Furthermore, I suspect that the number of cryptocurrencies in circulation will decline substantially from 20,000 to just a handful which will further reduce supply.
• Coinbase is another example that, assuming that it has not lied to its auditors, has a solid balance sheet and is managing client assets in a prudent and reasonable fashion.
• I still think that the shares of Coinbase are too expensive for what it is but crucially, I think it survives the winter and will be part of whatever arises from the ashes of this bust.
• I think Binance is another survivor although I am unable to verify its balance sheet as it is privately held.
• The main problem with crypto today is that it is really bad at its job which I think is caused by its technology not being very mature.
• For example, cryptocurrencies are nothing more than speculative assets as they make awful mediums of exchange or payment rails.
• A payment system needs to be fast, flexible, cheap and the token of exchange needs to be stable, and cryptocurrencies meet none of these criteria.
• This is why RFM Research has concluded that cryptocurrencies are wholly unsuited to run the economy of the Metaverse or the Internet leaving systems like WeChatPay and AliPay as far superior options.
• Non-fungible tokens (NFTs) are also wholly unsuited for the ownership of digital assets (like artworks) as only the token itself is on the blockchain while the asset sits somewhere else usually on the end of an HTML link.
• This is due to the very high cost of holding large digital assets on the blockchain.
• However, when these problems are solved, then blockchain will be in a position to fulfil its potential.
• This is the point that I suspect blockchain begins its long march back to credibility and investibility.
• I do not think that this will be accompanied by a rally in cryptocurrencies but is more likely to be in the companies that use the blockchain to provide services.
• On this basis, I expect that both Binance and Coinbase will reinvent themselves as I don’t think that there is a viable future in facilitating wild speculation in cryptocurrencies.
• This crash looks exactly like the internet bubble of 2000 and the real estate bubble in 2008 both of which saw the weaklings weeded out and the strong survive and thrive.
• I have long held the opinion that it is not until cryptocurrencies become really boring that they become viable and the speed at which the world is losing interest is a good sign that this journey has begun.
• I am not even close to wanting to pick up the pieces of this yet but with only a few of the majors left in businesses, the bottom feels like it is close.
• This may be followed by a long winter while the blockchain gets its house in order so that it can become a proper payment rail and a proper system for ownership of digital and intangible assets.
• Once, this has been achieved, then the stage will be set for a long-term, fundamentally driven recovery in the world of crypto.
• One to keep an eye on.

• The collapse of FTX is doing further damage to the now very fragile crypto proposition but I have no doubt that once the dust settles, blockchain & cryptocurrencies will survive as will Coinbase although I suspect that the shares still have further to fall.
• However, FTX is almost certainly worthless in yet another example of Softbank leading the VC lambs to the slaughter.
• FTX is a crypto exchange where investors can trade and hold crypto assets but it appears that its finances were so fragile that a sentiment shift caused by a series of tweets may have been its undoing.
• It seems that a fall in the price of FTX’s issued FTT coin has triggered a loss of confidence in FTX which has led to a run on its assets which in turn has triggered an urgent call for vast amounts of new money to pay back depositors.
• This is where the questions begin because as an exchange, FTX should not have meaningful exposure to cryptocurrency assets.
• Instead, it takes customer deposits, facilitates trading and then provides custody services for the crypto assets that its customers choose to invest in.
• Consequently, when investors ask for their money back it should be able to provide all of that cash in very short order to its customers.
• It does not have a bank’s classic duration mismatch where short-term deposits are lent out on a long-term basis meaning that it is unable to pay back all of the deposits in one go.
• Hence, a “run on the bank” should not cause this kind of distress in exactly the same way that Coinbase was able to pay back $3.5bn to its depositors over the last 6 months with barely a murmur.
• The fact that Sam Bankman-Fried (SBF) is desperately seeking between $4bn – $8bn to meet its withdrawal demands and prevent collapse raises all of the wrong sorts of questions.
• FTX is a privately held company and so there is no way of telling exactly what is on its balance sheet, but I am pretty sure that it is nothing like as clean as Coinbase’s.
• My guess would be that somehow its customers’ fiat currency deposits have ended up being exposed to the FTT coin and/or other cryptocurrencies meaning that FTX has no way to meet the demands for deposit returns.
• Furthermore, the company was already under investigation by the SEC aimed at looking at how FTX manages customer deposits.
• FTX’s investors now have some serious egg on their faces including Sequoia which claims it ran a “rigorous” due diligence process late in 2021.
• Despite this due diligence, Sequoia somehow concluded that the company was worth $25bn despite earning just $1bn in revenues and $250m in operating income in 2021.
• This sounds to me like more Softbank-like FOMO (fear of missing out) which is rapidly becoming a reliable contrarian indicator.
• Bitcoin has now broken the all-important $20,000 support level with the next stop being somewhere around $10,000.
• There is no end in sight to the continuing rout in crypto, but at some point, all of the hype, speculation and nonsense will have been flushed out, leaving the bare bones of the proposition.
• These barebones do have a use case such as fractional ownership of indivisible and intangible assets as well as running the economy of the Metaverse (once cryptocurrencies become very boring) and it is on this basis that I would look at this space.
• On this basis, I think Coinbase survives.
• As long as the company is not defrauding its auditors, the balance sheet is clean, it has only a small exposure to crypto itself and it can quickly pay back all of its depositors should it need to.
• This makes it a bank in my opinion but on that basis, it is still too expensive especially if its deposit base is going to decline heavily during the ongoing crash.
• The best US banks trade at 1.0x – 1.5x price to book value and Coinbase is still on around 2.0x.
• Hence, I think it has much further to fall but it will be one of the survivors of this death spiral.

• Cryptocurrencies have stabilised around current levels but activity in the space has declined greatly, meaning that exchanges and banks need to make the investment case on valuation which is something that Coinbase is still not able to do.
• Coinbase reported depressing Q3 22 results with revenues / EPS of $590.3m / LOSS$2.43 compared to forecasts of $641.8m / LOSS$2.38.
• In a sign of just how much things have worsened, revenue in Q3 2021 was $1.2bn and $803m in Q2 2022.
• The only bright spot was interest rates which drove interest income significantly higher as a result of the client assets that it holds in deeply liquid and secure US government securities where yields have been rising.
• Partnerships with Blackrock and Google also help establish Coinbase as the leader in its space, but the company remains very top-heavy in terms of expenses and much more action is needed.
• For example, R&D is currently running at 94% of revenues while administrative expenses are at 62% of revenues.
• For a technology company with these revenues, R&D should be around 20-30% (including the stock-based compensation) while admin should be about 5%.
• Coinbase has a very long way to go in terms of right-sizing the businesses and dealing with the hangover of the crazy crypto party that it enjoyed.
• If this can be achieved, then Coinbase becomes an interesting proposition as long as a valuation argument can be made for it.
• This is because, at its heart, Coinbase is a bank and not an asset manager or a trading platform.
• This is a key distinction because of the way that client assets are held by banks as opposed to asset managers.
• Asset managers and trading platforms are required to segregate client assets meaning that if the company goes bankrupt, the client assets are unaffected.
• Banks are different in that they borrow money from depositors, lend it out at a higher rate and keep the difference in interest rate between what they pay and what they earn.
• There can be no segregation of assets which is why if there is a run on the bank and everyone wants to get their money out at the same time, the bank will quickly become insolvent.
• Coinbase is set up as a bank except that it does not lend the money out but makes money on fees and services when its depositors trade in crypto assets.
• The company itself has a relatively small direct exposure to cryptocurrencies as client deposits that are not invested in cryptocurrencies are stored in deeply liquid and very secure money market instruments.
• These deposits are kept in financial institutions and are covered by the usual depositor insurance.
• This means that Coinbase has not lent its clients’ money out as term loans meaning that if everyone suddenly wanted their money back, Coinbase would be able to oblige them although it might take a week or two.
• This has happened to some degree as client assets have fallen significantly which Coinbase has been able to meet without any hiccups.
• Furthermore, the chance of the company going bankrupt is very very low even if there is a total crypto meltdown and the sector goes into a nuclear winter.
• This is because the company still has $5bn on its balance sheet and $3.4bn in debt giving a net cash position of $1.6bn.
• Client assets have fallen to $6.5nm from $10bn around 6 months ago and this will be the key factor to watch going forward.
• Client assets need to stop declining and stabilise before anyone can have any real idea where to right-size this business in terms of expenses.
• Here there still remains a lot to do and I am not very comfortable that management is really committed to the deep and hard cuts that are now needed.
• This is because Coinbase has the classic technology company voting structure where Mr Armstrong, founder and CEO controls the company through a super-voting distribution of shares which carry 20x the number of votes that other shareholder hold.
• Like Mr Zuckerberg, if Mr Armstrong chooses to run the company into the ground, there is very little that anyone can do to stop him.
• This is what makes me nervous as the commentary in the results is not nearly aggressive enough when it comes to managing expenses which to me sounds like a lack of commitment.
• This is a shame because Coinbase has the potential to be a fantastic business as it has gross margins of 83%, is the leader in its field and does not have the classic duration mismatch on assets and liabilities that can cause banks to collapse when depositors demand their money back.
• It is also a shame because the price-to-book ratio (the classic way to value a bank) is 2.2x which compares very unfavourably with the best quality US banks which trade at 1.0x and 1.5x.
• Hence, I think Coinbase has much further to fall and it will need to be a lot cheaper before I would consider it over the classic banking sector.
• This is another one of those to watch as a further deepening of the current crypto crash is likely to push what is a pretty well-capitalised bank into deeply discounted, highly attractive territory.

• Crypto miners that were rendered homeless by the Merge are struggling to find economic uses for their equipment meaning that they are very likely to sell up and leave the industry.
• On Thursday, September 15^th, the mechanism by which the Ethereum blockchain is managed changed from computers solving complex problems (proof of work) to token owners pledging their tokens as collateral to guarantee transactions (proof of stake).
• The main result of this is that the energy consumption of Ethereum fell from being equivalent to all of Chile to almost zero but it also meant that those that were doing the calculations were out of a job.
• To date, these calculations were carried out by miners using graphics cards which are ideally suited to this task but there is now around $10bn of equipment now looking for something to do.
• There are other coins that can be mined such as Ravencoin and Ergo but those that have tried have quickly found that the venture is loss-making just on the cost of electricity alone.
• It should be possible to eke out a small margin in territories where the cost of electricity is lower but even then, the margin is likely to be mere cents on the dollar
• Furthermore, this is before one takes into account the cost of the equipment which I suspect will push almost all of the miners into loss-making territory even where energy is cheaper.
• Finally, a significant slice of the equipment has been financed through borrowing and now that revenues have collapsed, these miners are likely to be forced sellers.
• Some miners are holding on in the hope that the crypto market turns around meaning that they can become profitable again but at least half have already switched off their equipment.
• This gives weight to my view that around $3bn – $5bn of graphics cards are about to hit the 2^nd hand market which is going to depress the price of graphics cards substantially.
• It is already possible to find high-end Nvidia cards on Amazon selling at a price that is 60% below where it was just 12 months ago.
• This is bad news for AMD and especially Nvidia but the ramifications for Nvidia are far less than for the miners as its Cuda core still remains ideal for graphics for gaming, AI in the datacentre, the Metaverse as well as a number of other areas.
• I can see another couple more bad quarters for Nvidia as gaming performs worse than expected as gamers will be required to absorb graphics cards being sold by obsolete miners into the second-hand market.
• It is important to note that as of FQ2 2023, the data centre is twice the size of the gaming segment meaning that its ability to hurt Nvidia’s overall performance has already been significantly curtailed.
• I still remain extremely cautious on cryptocurrency which shows no real sign of emerging from its current deep freeze and in semiconductors I continue to prefer the cheaper, more defensive end of the sector.
• Qualcomm, MediaTek and TSMC all offer good growth at valuations which are much cheaper than AMD or Nvidia.
• I would like to own Nvidia one day, but I suspect that the crypto crunch may give me the opportunity to buy it at levels below even here.

• Ethermine, the world’s biggest mining services provider for Ethereum miners will close its servers over the next 24 hours rendering $10bn of hardware useless which is now likely to end up on the second-hand market.
• While this is a big step forward for cryptocurrencies in their quest to become viable payment mechanisms (see here), it is likely to create a larger hardware overhang.
• It is estimated that there are around 1m Ethereum miners who have around $10bn of hardware between them, a large piece of which is going to be graphics cards.
• This is because crypto mining is much better suited to the kind of processors that are used for graphics and I suspect that there are around $3bn – $5bn worth of graphics cards that will now be looking for a new home.
• This is in addition to the significant increase in supply that resulted from the collapse in Bitcoin valuation which combined with increasing energy prices, rendered many miners uneconomical.
• Nvidia has already seen the impact of this on its gaming business in its last quarterly results (see here).
• This is more bad news for Nvidia and AMD, but this is a dent rather than a crater for Nvidia.
• This is because its graphics cores are already being used in many other applications and so after a couple of difficult gaming quarters, I expect this cloud will pass.

Twitter – Obvious approval

• Twitter shareholders have made the obvious move in approving Elon Musk’s takeover of Twitter but forcing him to complete the transaction will be quite another matter entirely.
• The transaction was approved in a short conference call in a demonstration of just what a good deal Mr Musk was offering.
• This is further demonstrated by the bad case of buyer’s remorse from which he is now suffering.
• This is leading to him pulling every manoeuvre that he can think of to try and get out of the commitment that he made.
• This is why he recently latched onto the testimony of Pieter Zatko who testified to congress on a national security issue.
• His testimony tends to support Mr Musk’s reason for abandoning the deal which is that Twitter has been less than truthful with regard to the fake and spam accounts that it has on its platform and its general level of data security.
• However, given that Mr Zatko was fired by Twitter in January, it is clear that he has an axe to grind with the company.
• Hence, I continue to think that the Delaware court will order Mr Musk to buy the company but that it will prove almost impossible to enforce (see here).
• Hence, I suspect that the deal will not go through and given what has happened to the rest of the technology sector, it is not hard to argue that Twitter should be trading at $17 – $20 per share.
• Hence, Twitter shareholders waiting for $54.20 per share are unlikely to realise that price and given the downside that still exists, are best selling out now as they are still getting a great deal for their shares.

• The Ethereum Blockchain will attempt to change the way it operates this week which will make it much cheaper to operate (good for payments) but at the same time, there will be a load of graphics cards that suddenly are no longer needed by their owners (bad for Nvidia).
• There are two big blockchains in use today, the Bitcoin blockchain which has a market capitalisation of $370bn and the Ethereum blockchain which has a market capitalisation of $201bn.
• However, a key difference is that the Ethereum blockchain does far more than the Ether cryptocurrency that competes with Bitcoin as it also runs most NFT systems as well as decentralised finance.
• Consequently, the Ethereum blockchain is larger than it looks and in my opinion is probably more important to the crypto industry than the Bitcoin blockchain.
• While the blockchain systems are decentralised and independent of any centralised control they also suffer from being expensive to transact, inflexible, slow and unstable in terms of monetary value.
• This makes crypto wholly unsuitable for being used as a payment system or medium of exchange.
• This explodes the myth that crypto is needed to run the economy of the Metaverse as in its current state, systems like AliPay and WeChat Pay are far better suited (see here).
• Furthermore, blockchain transactions are verified by a system called proof-of-work which involves very computationally heavy mathematics which is where the enormous energy consumption of blockchains comes from.
• To put it in context, Ethereum consumes more power than Chile and produces more carbon than Finland while Bitcoin consumes more energy than Belgium.
• These calculations are best executed on Nvidia graphics chips which is one reason why Nvidia did so well when the crypto craze was in full swing.
• Gaming revenues increased by a factor of 3x between FY2017 and FY2022 to top $11bn.
• The Merge involves the Ethereum blockchain merging with another system that does not use proof-of-work but uses a system called proof-of-stake.
• Simply put, proof-of-stake means that those that hold Ether tokens pledge those tokens as collateral in order to validate the transaction.
• There are some potential shortcomings with this system, but because it no longer uses computation for verification, its energy consumption should fall by 99% or more.
• Given how much energy Ethereum consumes and how much carbon it produces, this is a very significant move and while it will not solve any of the problems of blockchain as a payment mechanism, I think it is a big step in the right direction.
• One of the reasons why transactions are expensive on the blockchain is due to the cost of verification which will now fall substantially should the Merge be successful.
• I see this as the first of many steps towards making crypto usable for transactions, but a lot still needs to be done if crypto wants to run the economy of the Metaverse or anything else.
• Furthermore, if the Merge is successful then I suspect that many of the other blockchains will begin to migrate over to proof-of-stake over time.
• This will be nothing short of a complete disaster for the picks and shovels of the crypto industry as demand for their products could effectively go to zero.
• The largest one of these is privately held Chinese mining equipment maker Bitmain but further down the supply chain are Nvidia and to a lesser extent, AMD.
• Nvidia’s unique system of massively parallel computing with its Cuda cores is ideal for executing the computations required for proof-of-work.
• If this moves to proof-of-stake then the bitcoin miners will no longer need the graphics cards they have bought and the increase of graphics cards we have seen coming into the second hand market as a result of the crypto winter could become a flood (see here).
• While this could easily put vendors like Bitmain out of business, the ramifications for Nvidia are far less as Cuda still remains ideal for graphics for gaming, AI in the datacentre, the Metaverse as well as a number of other areas.
• Nvidia has millions of developers across a wide range of applications and so while proof-of-stake may hurt Nvidia in the short term, it is a dent rather than a crater.
• Hence, if the Merge is a success, I can see another couple more bad quarters for Nvidia as gaming performs worse than expected as gamers will be required to absorb graphics cards being sold by obsolete miners into the second-hand market.
• It is important to note that as of FQ2 2023, the data centre is twice the size of the gaming segment meaning that its ability to hurt Nvidia’s overall performance has already been significantly curtailed.
• I still remain extremely cautious on cryptocurrency which shows no real sign of emerging from its current deep freeze and in semiconductors I continue to prefer the cheaper, more defensive end of the sector.
• Qualcomm, MediaTek and TSMC all offer good growth at valuations which are much cheaper than AMD or Nvidia.
• I would like to own Nvidia one day, but I suspect that the crypto crunch may give me the opportunity to buy it at levels below even here.

• Nvidia reported difficult results and soft guidance as the hangover from the crypto party continues to be a big headache.
• FQ2 2023 revenues / EPS were $6.7bn / $0.51 in line with revised guidance but well below its forecasts made three months ago (see here).
• As expected, the main culprit was Gaming where revenues fell by 44% QoQ and 33% YoY.
• Although data centre and automotive grew well, they could not overcome the sudden decline in Gaming, which forced Nvidia to warn so suddenly.
• The good news is that the 25.4% point hit to gross margins was caused by $1.22bn in charges for inventory, unused fab manufacturing capacity, provisions taken against weaker future demand and a $122 provision for warranty reserves.
• If I take these out, I end up with gross margins of 63.5% which is close to the levels that the company has recently enjoyed.
• Therefore, I expect that gross margins will bounce back in FQ3 2023 but there is going to be pressure from the sudden fall in demand which looks set to continue into FQ3 2023.
• Nvidia has taken this into account in its guidance where it is expecting FQ3 2023 revenues / gross margins of $5.9bn / 62.4% (+/- 50bp).
• This is below the already reduced FQ3 2023 revenue consensus of $6.92bn and the shares fell by 9.2% on 26^th August although this was doubtless exacerbated by the so-called “hawkish” speech from Federal Reserve chairman Jay Powell.
• The biggest problem is that Nvidia has no real idea how much of the demand for graphics cards it has enjoyed in the last 12 months has been due to the crypto craze which is now solidly in full-scale crash territory.
• This is because it sells its chips into the channel and not directly to end users and so it can’t tell how much of its silicon goes into mining rigs and how much into games machines.
• My suspicion for a long time has been that a big piece of this demand has come from mining rigs but many of the miners are now unable to make money thanks to the rising price of energy and falling crypto valuations.
• This combined with the crash in sentiment towards Bitcoin and its peers as well as some high-profile collapses has put crypto into a deep freeze and there is no knowing when it is going to emerge.
• Hence, a lot of miners are now looking to exit and are putting their graphics cards into the second-hand market which is depressing prices and demand for new cards.
• The poor macro outlook has also had an impact, but the suddenness of the correction as well as anecdotal data points from the crypto industry leads me to believe that the majority of this is due to the crypto crash.
• The silver lining here is that this means that the core markets of gaming, AI, data centre and automotive continue to fare reasonably well and that the long-term outlook for Nvidia remains largely unchanged.
• Hence, the investment case all comes down to valuation and unfortunately, Nvidia’s valuation remains too high compared to its peers.
• Assuming a FY2024 recovery to $40bn in revenues and $5.50 per share, the company is still trading on a FY2024 PER of 30.0x and an EV / Sales of 10.0x.
• In this environment and with the outlook that Nvidia has given, this is still too high and so I think the share price is going to continue to struggle to make headway.
• This is where the defensive nature of Nvidia’s competitors plays in their favour as Qualcomm and MediaTek are also seeing slower growth, but their valuations are far less demanding at around 14.0x 2023 PER and they are outperforming as a result.
• Hence, I am more than comfortable to stick with the defensive end of the semiconductor sector safe in the knowledge that the market is likely to give me an opportunity to get interested in Nvidia once again.

• The rhetoric around the safe haven and inflation hedging properties of crypto has gone eerily quiet as it transpires that cryptocurrencies are nothing more than speculative assets whose value is impossible to determine.
• The news gets worse and worse as Bitcoin is now around $21,000 and many of the crypto-related companies are cutting staff as they grit their teeth and get ready for deep winter in their sector.
• The latest bout of volatility was triggered by Celsius Network which froze its users’ assets citing “extraordinary market conditions” but which to me looks like a run-of-the-mill run on a bank.
• Celsius Network is a crypto lender where holders of cryptocurrencies can deposit their tokens and earn high yields by allowing Celsius to lend them out.
• In the era of negative real interest rates, this has been an attractive proposition as the bond market to any rational investor has been unattractive for quite some time.
• However, it is also a very risky one as the underlying value of cryptocurrencies is impossible to determine.
• As investors (speculators) get nervous and want to withdraw their tokens and sell them, Celsius is unable to meet the demand as it has lent the tokens out to earn yield in a replay of an ages-old banking problem.
• This has triggered a deepening loss of confidence in cryptocurrencies resulting in the sector now being worth less than $1tn compared to $3tn a few months ago.
• Bitcoin’s proponents point to its finite supply as a reason for its value, but there is no limit to the number of cryptocurrencies that can be created which has recently ballooned to around 20,000.
• This means that the supply of cryptocurrencies is infinite just like fiat currency and one only has to look at what has happened to Luna for evidence of what that means.
• I have little doubt that soon there will be far fewer cryptocurrencies around and that Bitcoin will be one of the survivors but the big question is: at what price?
• Furthermore, any notion of the safe-haven or inflation hedge has been blown out of the water with the evaporation of 2/3rds of the value of the sector coinciding with a 40-year high in inflation that is destroying people’s savings by around 8.5% per year.
• Of all of the crypto companies, I think that Coinbase is one of the most interesting because, at its heart, it is an unregulated bank that can easily withstand withdrawals by panicked users.
• However, this company is still trading at over 2x PBV compared to the best banks in the industry which typically trade at 1.0x to 1.5x.
• Hence, there is no valuation support for Coinbase and the fact that it is cutting staff clearly points to the fact that it is also not very optimistic with regard to its outlook.
• As yields rise elsewhere and liquidity evaporates, the crypto sector is going to look less and less attractive, meaning that this is not a blip but a full-on resumption of reality.
• Cryptocurrencies remain unable to function properly as mediums of exchange and, as such, are wholly unsuited to running any economies be they in The Metaverse or in El Salvador.
• When these issues are fixed, then cryptocurrencies will find their place in the world of finance, but this could take a while as this will require a lot of changes to be made to the blockchain protocol.
• Until then, cryptocurrencies will remain purely speculative that I think will prove to be even riskier than dabbling in junior mining exploration companies.
• It was Mark Twain who coined the phrase that a mine is a hole in the ground with a liar standing next to it.
• It is strangely ironic that the cryptocurrency currency creation process is referred to as mining.

• The current scare surrounding Coinbase is entirely valid in terms of proper risk disclosure but I agree with the CEO that even in a total meltdown, the company is very unlikely to go bust.
• Due to a new regulation from the SEC, Coinbase was required to add a risk warning to its filings that make it extremely clear exactly what Coinbase is and the risks that its users are taking by using its platform.
• This warning basically states that if Coinbase was to go bankrupt then client assets would be treated as unsecured creditors meaning that both the deposits and the tokens in the wallets could be lost as the secured creditors would be paid back first.
• This caused consternation among shareholders and triggered a further collapse in the shares which had already been clobbered by a bad set of results and horrendous market conditions.
• The simplest way to understand the risks attached to Coinbase is to view it as a bank and not an asset manager or asset trading platform.
• This is a key distinction because of the way that client assets are held by banks as opposed to asset managers.
• Asset managers and trading platforms are required to segregate client assets meaning that if the company goes bankrupt, then the client assets are unaffected.
• Banks are different in that they borrow money from depositors, lend it out at a higher rate and keep the difference in interest rate between what they pay and what they earn.
• There can be no segregation of assets which is why if there is a run on the bank and everyone wants to get their money out at the same time, the bank will quickly become insolvent.
• Coinbase is a set up as a bank except that it does not lend the money out but makes money on fees and services when its depositors trade in crypto assets.
• The company itself has a relatively small direct exposure to cryptocurrencies as client deposits that are not invested in cryptocurrencies are stored in deeply liquid and very secure money market instruments.
• These deposits are kept in financial institutions and are covered by the usual depositor insurance.
• This means that Coinbase has not lent its clients’ money out as term loans meaning that if everyone suddenly wanted their money back, Coinbase would be able to oblige them although it might take a week or two.
• Furthermore, the chance of the company going bankrupt is very very low even if there is a total crypto meltdown and the sector goes into a nuclear winter.
• This is due to the strength of the company’s balance sheet which I suspect that almost no one has looked at.
• As of March 31^st 2022, the company had $6.1bn of cash and $3.4bn of debt giving the company a $2.7bn net cash position.
• It also is holding $9.7bn in custodial funds from customers (deposits) which had a market value of $10.0bn (highly liquid, money market instruments).
• Coinbase holds $1.3bn in crypto assets for its own account and $246bn of crypto assets that belong to its clients, but the client assets are currently off-balance sheet.
• Part of the new regulation will require Coinbase to recognise a liability with regard to its obligation to safeguard client assets as well as an associated asset.
• This safeguard does not mean protecting clients from bad investments that they make, but to ensure that it has adequate safeguards to protect those assets from hacks and other preventable risks.
• Hence if the $246bn goes to zero, this is not on Coinbase and its trading losses would be limited to the $1.3bn that it holds for its own account.
• I think the company is strong enough to survive this eventuality given its net cash position.
• In Q1 2022, the company had gross margins of 79% but still lost $554m at the operating level due to its very high investments in technology and general and administrative expenses which have grown substantially over the last 12 months.
• If there is a nuclear winter for crypto, then Coinbase can drastically cut its OPEX to match the demand that it has for its services and probably even turn a profit and generate cash although this would be miles below where current forecasts are.
• Forecasts are on the way down given how badly the company missed its Q1 2022 revenue and the weakness of the guidance that it gave for Q2 2022.
• Coinbase shares are down 89.3% from the close of the first day that the company was publicly listed and so the question on everyone’s mind is where is the bottom?
• Coinbase’s business and risk profile look to me to be like a bank rather than an asset manager and as such, one should look at it as a bank.
• The best simple measure for bank valuation is price to book value and the biggest and best US banks currently trade between 1.0x and 1.5x.
• Coinbase is a low-risk bank but is active in a market that can only be described as high risk and speculative.
• Even after falling nearly 90%, it still trades at 2.3x PBV which leads me to think that there is plenty of room for it to fall further.
• I think it quite likely that Coinbase may sell off way too far before it bottoms but on the measures that I would use to look at this, that bottom is still far off.
• This is another one of those to watch as a deepening of the current crypto crash is likely to push what is a pretty well-capitalised bank into deeply discounted, highly attractive territory.

• The first signs of Intel’s new identity are beginning to emerge with the launch of a new product that aims to challenge Chinese dominance of equipment for Bitcoin mining / Blockchain computing.
• This is something that is badly needed in this industry where the verification of transactions on the distributed ledger is inflexible, slow and expensive, all of which make the blockchain unsuitable for being the backbone of a digital economy.
• The Bitcoin mining industry is currently dominated by Bitmain and MicroBT who sell the hardware required to mine Bitcoin, but the disruption caused by Beijing’s clampdown has created an opportunity for competition to come in.
• This industry is all about energy efficiency as, after the capital cost of the equipment, the only real cost incurred by the miners is their electricity bill.
• Here, the Chinese have designed custom ASICs (application-specific integrated circuits) whose sole purpose is to run the Bitcoin mining algorithm as efficiently as possible.
• As a result, equipment from Bitmain and MicroBT lead the industry in terms of energy consumption which is why they are currently the market leaders.
• This is measured in joules (unit of energy) per terahash (unit of compute calculations) where Bitmain claims 21.5 and MicroBT claims 31.0 far better than the competition (including Intel) who are currently in the mid-fifties.
• This means that Intel has a lot of ground to make up but there are some other areas where it can make an immediate difference.
• The biggest of these is fixed pricing.
• Bitmain and MicroBT change the prices of the machines that customers have already ordered but not yet received due to the volatility of the Bitcoin price, which greatly complicates life for the miners.
• Intel is not doing this and is sticking to its normal practice of charging the prevailing price at the time at which the product is ordered.
• This will go some way towards offsetting Intel’s disadvantage in joules / terahash as future visibility has a significant economic value.
• I think that Intel is playing the long game here.
• The real future of this product line is not in Bitcoin mining but in the maintenance of the blockchain distributed ledger more generally.
• Blockchain has been touted as the backbone of the economy of the Metaverse, but its cost, speed and flexibility currently render it completely unsuitable to carry out this function.
• RFM research has found that there is no fundamental reason why the traditional payment rails cannot be used to run the economy of the Metaverse meaning that the blockchain has a hill to climb.
• Intel’s long-term expertise in designing and manufacturing semiconductors as well as its substantial resources mean that it is in a position to provide substantial competitive pressure.
• This is exactly what the blockchain needs as this pressure will accelerate competition and bring forward the time when the blockchain is in a position to run a digital economy.
• This combined with a return to manufacturing leadership is what I need to see from Intel, but substantial challenges remain.
• There is a significant risk that the x86 architecture upon which almost all of Intel’s chips are based is now obsolete and Intel needs to emphatically demonstrate that this is not the case.
• Its competitors (Samsung and TSMC) are spending more on Capex leading to the possibility that it is unable to catch up and must now resign itself to the second tier of an industry it once dominated.
• Furthermore, the financial outlook for the medium term has deteriorated exactly when it should be booming given the shortages, the semiconductor super cycle and the strong results of its peers.
• EPS for 2022 is now expected to be in the range of $3.50 which gives a 2022 PER of 12.7x which is not expensive but not exactly cheap when compared to its peers and the challenges the company currently faces.
• Intel is knocking on the door of value territory but while its earnings continue to deteriorate and the uncertainty persists, its value is also deteriorating.
• I have previously floated the idea of Intel at $44 but at that time, that would have represented a 12-month forward PER ratio of around 9.5x.
• This number is now closer to 13x which I think does not compare very favourably to Qualcomm which is still on around 15x.
• Qualcomm has none of the strategic challenges that Intel faces, is growing faster and its earnings are only 15% more expensive.
• Hence, I think Intel shares can go lower as the value argument for Intel is still too elusive to put real money behind.
• I think the shares go lower or continue to underperform their peers.