With all sorts of interoperability solutions being implemented, we can now access the "tech" of one cryptocurrency with the token of another (i.e. wrapped bitcoin on ethereum.) Which features aren't trivial to access cross chain? /1
2/ I frequently point out the time gap - and this matters. We may eventually get a ZK-rollup chain on bitcoin for competitive privacy for example, but if it takes too long, that creates room for competition to gain a foothold. How long is too long? Don't know, but many years.
3/ but let's assume for now that interoperability solutions will come swiftly and render the tech competition superficially trivial. I.e. Can access any tech from any chain, but possibly with meaningful friction for specific use cases like defi. What will determine winners?
High level thoughts on what I think happens over the next 5 years in cryptocurrency (not what I think should happen, or want to happen, but my guess as to what will.) I think periods of rising bitcoin dominance will be countertrend and temporary. /1
2/ at top of this bull market (maybe Q1-Q2 next year?) I think BTC likely to be the 2nd or even 3rd largest coin by market cap and <4th in network fees and usage. The work to add critical functionality to bitcoin to natively support basic financial features likely
3/ requires the introduction of new opcodes and then a bunch of work building financial primitive libraries so people can build dApps. IMO, the solidity sidechain based dApps on bitcoin will get usage, but won't have a clear enough consensus to feel like "the future."
*A crypto education resource thread* (please add in helpful links). Let's start with this excellent compendium of resources of all types: blockbyblock.io/resources /1
2/ for news I typically check messari, theblockcrypto, and then rely on crypto twitter and telegram. For watching markets, can't beat tradingview. For market data: skew, bybt, and genesisvolatility. For on-chain data, glassnode and cryptoquant.
Have thought a lot about this. There's an interesting dynamic where the phenomena as a whole is "stupid" and investors in aggregate will lose money, yet it's totally rational for "smart money" to participate in inflating the bubble. /1
2/ in really brief form - most people realize there's a bubble and that prices will be lower in the future, but everyone is hoping/expecting to be able to play the game better than others and exit before the music stops. In aggregate, wrong, but individually possible.
3/ and there's immense pressure to "play the game." As Citibank's former CEO said, "As long as the music is playing, you’ve got to get up and dance." Fund managers and CEOs are expected to perform on a quarterly basis. If they "sit out" a bubble, typically long before
Thoughts on hedging in crypto: Tough to do for anything beyond short time frames. A. Derivatives rarely offer "cheap" hedges. B. Correlations are non-stationary. B. Most hedges require taking on additional severe counterparty risk. But there are some choices. /1
2/ A. First on derivatives - the only derivatives with real liquidity in crypto are BTC and ETH vanilla options (calls and puts). Implied volatility over the past year has occasionally gotten moderately cheap after periods of quiet. Could buy 1 or 3 month puts as a hedge.
3/ certainly perpetual futures and futures generally are also liquid but aren't really "hedges" for most crypto investors. If you're long ETH and sell an ETH future for example, yes, you're now "hedged" but you've cancelled your investment and transformed it into a basis trade.
A simple educational thread on options below. For the more sophisticated practitioners, here's a new thread on a concept rarely talked about and poorly understood even amongst professional market makers: serial autocorrelation and its impact on option pricing. /1
2/ Initially, options were priced using Black-Scholes formula, which assumes markets follow a "random walk" - that they exhibit no predictable momentum or mean reversion (and also assumes zero skew and 'normal' kurtosis of 3). This simplistic model is pretty far from reality.
3/ By the early 1990s, options were consistently being priced to incorporate skew and kurtosis. This thread isn't about those terms, but simplistically - those are measures of the 'tails' of the distribution.
Want to get rich in crypto? Unless you have serious trading chops and $1m+ in capital, your best bet is to join a crypto firm to gain a network/paycheck/education. Then you can patiently invest/trade on the side and spot great opportunities to join/start young projects. /1
2/ which firm should you join? Not as important as you might think, as long as you can get general networking and educational exposure from there. Big firms or prominent project teams (e.g. Coinbase or Uniswap) are the safest launching pads.
3/ not everyone can get hired there of course, so it’s fine to accept a decent but smaller project/company which can always be a stepping stone to greater things if necessary.
An interesting thread. The one thing I'd disagree with: the cool stuff that comes with being rich doesn't all happen automatically. If you don't consciously build a network of the right service providers, you're likely to just get taken advantage of. /1
2/ it requires a little skill and work (or at least a a bit of effort) to convert money into the things mentioned in that thread. An example in my personal life was when I was recently stuck in a costa rican jungle in the middle of the night.
3/ I was able to hike to get cell reception, but had no obvious way to convert even an infinite amount of money into a fast rescue. It took some elbow grease, negotiation, and $1k in gratuities. But $100m wouldn't have sped things up by itself.
What's the best place to go for practical discussions of PoS game theory (particularly through lens of ethereum's upcoming transition)? I'm a bit out of date on the topic. Tell me why this is wrong: the cost of attack is likely marginally higher than prevailing lending rates /1
2/ an attacker needs to borrow x% of eth for y amount of time. The interest cost of that borrow over that period of time is the cost of attack. The one "bottleneck" is locating sufficient loans.
3/ if this is correct, than the interest rate probably doesn't matter. Unlike with buying (where you'd drive the price up tremendously acquiring a large stack), loan rates are more homogenous such that you likely wouldn't have to drive rates up all that much to attract nearly all
ESG in practice is similar to when companies brand products as green, or add a cure cancer symbol and donate some trivial % of sales. It’s marketing with very little substance. Both consumers and investors want to feel good about themselves without sacrificing price or returns.
2/ Why so much marketing effort (and sales success) around such little substance? I think it's a chain effect of an "intolerant minority", and shows both the strength and weakness of that "intolerance." A small % of investors (and consumers) care about these topics enough to
3/ at least somewhat affect their behavior. They go to their pension, endowment, fund manager, or favorite retailer, and demand that action to be taken. But they don't care enough to accept meaningfully lower returns or higher prices as long as the concerns are paid lip service.
Was asked an interesting question by an old crypto friend: he noted that I seem to assess some projects like mobilecoin very differently than others (in this case, Dash). In some cases I seem to care about the concentration of token ownership, but not in others. /1
2/ I assess companies and cryptocurrencies very differently depending on what they're trying to be and their target use case. For example, if you're valuing a local restaurant, you probably won't spend much time on their software or global brand.
3/ for a proof of stake network, the distribution of coin ownership is paramount, since that's what is literally securing the network. Ownership distribution is important but far less so for a PoW network.
Bayes theorem is the heart of a great deal of practical probability. Imo, it should be taught in late middle school, and statistics/probability should replace calculus as basic math education after arithmetic. /1
2/ I probably use calculus once a quarter in my work (which includes a good ideal of applied math, derivatives modeling, quantitative portfolio management and risk assessment etc.). I use Bayes theorem *daily* both in work and in day to day life.
3/ Bayes theorem comes up constantly. I use it when I’m evaluating which vaccine to get, what hospital and medications to trust, interpreting political polls and policies, and even whether to take an umbrella to work.
"Every musician wants to be a comedian." <- a profound recruiting insight that I've learned the hard way. A thread. /1
2/ the quote references the idea that many musicians enjoy chatting up an audience in between sets. For some, it's not enough to be great at what they do (musician), they want to be involved in many things (general entertainment). It's the generalist vs specialist dilemma.
3/ Founders are often generalists, and generalists play a critical role in many organizations helping to synthesize and "see the forest for the trees." But an organization of nothing but generalists inevitably fails. Excellence in most areas requires specialization.
Crypto currently offers basically no economies of scale. I.e. with the same skills, a professional trader/investor will earn a far, far higher ROI running a smaller amount of their own money than launching a fund. /1
2/ fund management also brings it with expenses and headaches on the operational, legal, compliance, accounting, HR etc side of things. Join a fund if you want to learn, gain access to tools/platform, and moderately diversify economic exposure.
3/ launch a fund if you have the team/skills necessary (not just investing but ops, marketing etc), and want to spend 5+ years in an entrepreneurial grind to build a company.
Paradox #15: Braess's paradox - adding roads may slow down traffic. This is really two separate paradoxes. /1
2/ A. empirically, people seem to target a specific work commute time. If you relieve traffic congestion, people move further away to retain the same commute time, likely because the greater distance gives them a higher quality of life in other ways, like cheaper real estate.
3/ so you could say that adding roads doesn't do anything...but in this framing it does. It may not reduce commute times, but it's still adding value to peoples' lives. Effectively those extra roads let people get cheaper rent or prettier home surroundings.
The scope of regulation over anything and everything is mindboggling. 2 years ago, BlockTower held an analyst competition where we gave away $30k of BTC in prizes for the best investment analysis submitted. We had to review state by state regulation on "lotteries", and /1
2/ prohibit participation from many countries and states. This kind of thing comes up everywhere. Want to sell an NFT? Depending on how you're doing it, it falls under dozens of local and federal regulations (mostly irrelevant to individuals.)
3/ we're going to continue seeing the centralized/decentralized bifurcation as a regulatory arbitrage. The cost savings of being an "anonymous" founding team is enormous. (worth noting though, anonymity rarely lasts). Doing anything as a regulated business basically requires
Are NFTs securities? No, by default. In some specific cases, yes, depending on what they are and how they're marketed. (I'm not a lawyer, may be wrong.) /1 itsartlaw.org/2019/11/19/fra….
2/ Remember the howey test? Same test applies to NFTs. Does an artist selling a work of digital art in NFT form constitute a security? Usually no. It's not an investment in a common enterprise, and generally it's not marketed with an expectation of profit.
3/ Can NFTs be securities? Sure. If a work of digital art is marketed as an investment for profit and if that profit expectation depends a third party promoter's efforts, and if the investment itself can be interpreted as in a common enterprise, than it may be a security.
Some thoughts on value accrual post-interoperability: we've seen many recent announcements of major defi dApps migrating or replicating themselves on all sorts of layer 1s and 2s. Will these naturally aggregate on a single shard/chain/rollup? Do layer1s benefit from activity?
2/ this comes back to the old "fat protocol" debates of 2017 and earlier. Does having a lot of, say, defi activity flowing through a layer 1 or 2 give the tokens of that layer value? Most assertions of "yes" rely on one of the following:
3/ fee burning (like EIP 1559) is one way of converting usage directly into platform token value. Thorchain has another direct approach with bonded rune value tied to TVL. Polkadot's parachain auctions are an indirect approach.
Paradox #14 (bringing back this series, starting with an easy one): The friendship paradox - "most peoples' friends have more friends than they do." /1
2/ This statement is literally true, by virtue of the fact that some people have more friends than others, and those people with more friends are more likely to be selected as a point of comparison. Consider the person with 1 friend - they'll appear in a survey as
3/ one person's friend, weighing down the average # of friends' friends, for only one person. In contrast, consider someone with 2,000 friends. That person's 2000 friends end up appearing 2,000 separate times in the survey. The people with more friends are more likely to
Quick thoughts on bitcoin risk and return - since late 2016, I've been "all-in crypto." I've had 90% of my investable assets in cryptocurrencies, bet my career on crypto in mid 2017. I basically just had enough outside of crypto (in USD and gold) to sleep easy. /1
2/ this was an extremely aggressive allocation reflecting my conviction that bitcoin and cryptocurrency more generally offered an exceptional and extremely rare asymmetric return profile. I thought a 100x+ was quite likely in 2016, making the risk of 85%+ losses tolerable.
3/ as time goes on, bitcoin and crypto generally offer weakening asymmetry imo. While BTC may still provide a 100x, for that to happen today, BTC would need to achieve a $100 trillion market cap, about 13x that of gold, and reflecting about 1/4 of total global wealth.
Some lessons from Black Thursday: 1. In the short-term, market structure matters more than anything else. Cascading liquidations of short-term exchange leverage controlled the price of BTC for 6+ hours.
2/ "arbitrages" often aren't. On Black Thursday, Bitmex traded $1k+ below Coinbase. Since Bitmex only accepted BTC as collateral, no amount of collateral protected from liquidation. As BTC fell, the collateral value fell, with no natural end (until Bitmex was taken offline).
3/ Crypto has segmented market participants. A lot of buyers react over 1-3 days to decide to allocate more fiat, wire fiat to exchange, then make a buy. A lack of buyers at one moment doesn't mean they're not there, they often just need time.