@Anthony @erik.bjareholt Okay, sounds good.

---

Not directly related to the question, but let me share the following, which touches on a few aspects of the cryptocurrency markets.

Recent cryptocurrency distributions.

Two cryptocurrencies named Byteball (unit: GB; not to be confused with Bytecoin) and Stellar Lumens (unit: XLM) are both recently distributing some of their coins to Bitcoin (unit: BTC) holders based on snapshots of the Bitcoin blockchain at pre-specified times (action is needed to receive the coins).

Byteball just held their 6th round of distribution at 2017-06-09 13:10 UTC, where they gave 0.0625 GB for each BTC and 0.2 GB for each GB. Since each GB became 1.2 GB after the distribution, we should expect the market value of GB to have changed by 1/1.2 − 1 = −16.67% afterward, which we can check. We will consider the price of GB in BTC instead of USD since cryptocurrencies tend to have correlated USD prices. Using coinmarketcap.com, I find prices of 0.29847000 BTC/GB on the day before (2017-06-08 13:10 UTC) and 0.24900600 BTC/GB on the day after (2017-06-10 13:10 UTC), which gives a change of 0.24900600/0.29847000 − 1 = −16.57%. Not bad at all! (But we shouldn't put stock into this uncanny precision; treat this as we would −15% or −19% because going plus-or-minus a couple percent in 2 days would be well within usual volatility.)

The value of this distribution was non-negligible, since the 0.0625 GB given for each BTC equates to a 0.24900600 BTC/GB × 0.0625 GB/BTC = 1.56% return over just a day or so! The fact that only about 3% of BTC (450,000 BTC out of the 16,000,000 BTC total supply) were linked (thus receiving the distribution) suggests that many holders were unaware of this opportunity.

Stellar Lumens is about to have a distribution round on 2017-06-26, where each BTC receives about 975 XLM. At current market prices, this gives about 975 XLM/BTC × 0.00001568 BTC/XLM = 1.53% expected returns, over less than a day. We would not expect a significant drop in market price here because (unlike GB) each XLM before and after is pretty much the same, but I expect increased volatility.

The 7th round of GB distribution will be on 2017-07-09 (a month after the 6th distribution), where once again each BTC receives 0.0625 GB and each GB receives 0.2 GB. Assuming the current market price of 0.23774473 BTC/GB, I calculate 1.24% expected returns for this distribution (this figure takes into account the −16.67% change in GB price that I expect).

Participating in all three distributions (1.56%, 1.53%, 1.24%) would give about 4.33% returns over just a month. This is an APY of $\sf 1.0433^{12} - 1 = 66\%$ !

The return figures above were in BTC, not USD, so anyone who tries to receive these returns by temporarily exchanging USD to BTC will have exchange risks from Bitcoin's volatility in the few hours they need (to make sure they have control of the BTC during the blockchain snapshot) before exchanging them back to USD. (Theoretically, the price of BTC should drop by 1% to 2% soon after each of these distributions, just as stocks do after paying out dividends. However, looking back at the recent Byteball distribution, I don't see this on any of the historical charts from coinmarketcap.com and the GDAX and Bitfinex exchanges, which is disappointing. For those who want to check: I think the theoretical sell-off should have began between 3 blocks to 12 blocks after the snapshot at block 470514, meaning between 14:10 and 14:51.)

Neglected opportunity?

In any case, I reckon it's possible to receive these distributions while fully hedging against the price of BTC in USD; I think there might have been an under-recognized and under-used opportunity: We simply need to borrow BTC to receive these distributions, but the glaring question is: where can we borrow with sufficiently low interest rates and fees? Some exchanges allow margin lending; the Bitfinex exchange allows (non-US) customers to margin trade with 3.33x leverage on BTC/USD through automated peer-to-peer lending (they and similar exchanges do margin calls and forced liquidations on behalf of lenders to lower the risk of default; AFAIK no default has happened yet, but the risk is always there), but obviously they don't allow withdrawing borrowed BTC.

The simple solution I see is to buy BTC on a non-margin account, and simultaneously short the same amount of BTC on a margin account. By simultaneously buying and selling the same amount of BTC, we have fully hedged against the price of BTC in USD. Now we can simply withdraw the BTC we bought. There's no trickery and no free lunch going on here: we are effectively taking a secured loan of withdrawable BTC with 130% collateral and with interest fees; if we didn't want to withdraw, we wouldn't have had to put in that extra 100% collateral over 30%. (Note the risk of a margin call; more on this later.)

To have zero price slippage, we should execute the buy and sell orders against each other by simultaneously placing them at the same price within the bid–ask spread. I'm not sure if Bitfinex supports this on a single account, but the buy and sell can be placed on separate accounts if necessary. (Using two accounts would bring into question whether this is within the exchange's terms and conditions. In any case, if someone actually did this with a lot of capital, I think it would have been prudent to let Bitfinex's support know the details so that their systems don't freeze the account due to the trades and withdrawals, and so that they can clarify whether this is all within their T&C; none of these details have to be kept secret.)

Fee calculations.

Like I calculated before, I think these distributions give about a 4.33% expected return over a month. Let's round that down to 4% to be conservative. Furthermore, we need to divide that by 130% because margin borrowing on Bitfinex requires a 30% initial margin, giving a 4% / 130% = 3.08% return. But ignore this correction for now.

What about fees? We need to deposit and withdraw both USD and BTC, buy and sell BTC, and borrow BTC on margin. Let's list the fees if use Bitfinex:

Deposit USD at the start, and withdraw USD at the end: There's a 0.1% fee for in/out bank wire transfers and we need to transfer all 130% of our collateral, so the fee here is 2 × 0.1% × 130% = 0.26%.
Withdraw and deposit our borrowed BTC: Withdrawal costs 0.0001 BTC (≈ $0.3), and although Bitfinex doesn't charge for BTC deposits, we need to pay the network transaction fee of about 0.001 BTC (≈ $3). This adds up to a flat rate of about $10 if we deposit and withdraw 3 times each, which is under 0.01% if we transfer more than $100,000, so let's just call this 0%.
Buy and sell BTC: We need to buy and sell BTC 2 time each, and the Bitfinex fees for makers and takers are 0.1% and 0.2%, so as long as we match our orders against each other, we will pay 2 × (0.1% + 0.2%) × 100% = 0.60%.
Margin borrowing the BTC: Bitfinex charges only the lender, not the borrower, so we only need to pay interest. The market rate over the past few days is a little under 0.02% per day (= 7.57% APY), so over a month that's $\sf (1 + 0.02\%)^{30} - 1 = 0.60\%$ .

Remember I said to temporarily ignore the 130% correction? That's because we these fee calculations, like the profit calculations, are relative to the original 100%, so we need to add them up first and only then divide by 130%. This gives a net return of (4% − 0.26% − 0.60% − 0.60%) / 130% = 1.95%. Since that's over a month, this is an APY of $\sf (1 + 1.95\%)^{12} - 1 = 26\%$ .

Is a 1.95%/month return worth it? Granted, the average 1-year trailing return for stock index ETFs is currently around 18% = 1.39%/month, so we're getting just an extra 0.56%, while taking a lot of risk by trading cryptocurrencies, right? Not so. I reckon that the risk here is lower than that of stocks.

Managing risks.

Let's list our risks:

Recall that we have fully hedged against the price of bitcoin in USD, so there is no exchange risk from volatility, which I think is the dominant risk associated with trading cryptocurrencies.
Human error is a significant risk because this plan needs to be accurately executed to avoid accidental losses, but I think this risk is negligible given sufficient experience (not me).
A third risk is from losses due to the exchange, whether by breach or by misappropriation; this is a real risk given that Bitfinex was hacked in August 2016 and lost 36%(!) of user funds; they passed this loss to every customer's net balance. They managed to repay customers the entire 36% in April 2017, but they can't pay back the opportunity cost and risks remain in the future. IMO, the risk here is sufficiently small, but your mileage may vary. On the bright side, we've withdrawn the BTC we borrowed, so our maximum loss is the net value of our margin account, which, if the price of BTC stays the same, is 30% / 130% = 23.08% of the capital we began with.
The remaining risk that I see is that of a margin call if the price of BTC rises, potentially leading to forced liquidation of our margin position. Fortunately, we can easily manage this by closing part of our short position (i.e., buying BTC) while simultaneously selling the same amount of borrowed BTC, matching them against each other, so that we keep the value of the borrowed BTC at their initial value.

From my point of view, the combined risk here is lower than that of a stock index.

Can we actually execute these trades with a lot of capital, though? Given the effort needed, we'd want more than, say, $10,000 profit after the month-long period. To achieve this, we'd need at least $10,000 / 1.95% = $513,000 in capital. Let's round this up to $1 million and see if Bitfinex can handle this:

Can we easily trade $1 million? Bitfinex offers fee discounts for high-volume traders, and their highest level of discount is given to traders who trade at least $1 million per day on average over 30 consecutive days. So it looks like trading $1 million wouldn't be an issue.
Can we easily deposit and withdraw $1 million? Through wire transfer, I don't see why not.
Can we easily deposit and withdraw $1 million worth of BTC? Again, I don't see why not. It'd probably be a good idea to inform Bitfinex beforehand so that their automated systems don't unexpectedly freeze the account.
Can we borrow $1 million worth of BTC at low interest rates for a month? Right now, I see over 8000 BTC available for immediate lending with interest rate at or below 0.02%/day. This is over 8000 BTC × $2900/BTC = $23 million worth of BTC, of which some proportion guarantees this rate for 30 days or more (most of the loans guarantee their rates for only 2 day). I think it's likely at least $1 million of these BTC are available as 30-day loans (doesn't seem possible to check for sure). In any case, we can even borrow all of that $23 million if we take the risk that the interest rates might go up (if the rates go up, we can always close parts of our position until our maximum rate is low).

All in all, I think this profit opportunity was real and available. However, few people would be in a position to take it: for instance, at least for Bitfinex, they cannot be US residents, and they need to be able to liquidate at least around $100,000 for the needed cash without incurring large fees or taxes.

So is there anyone who can take advantage of this opportunity?

Yes, I think there is a definite set of people who can profit from the plan here. In fact, they can still do so. I chose to do the calculations over a month because I don't think it would be worthwhile for anyone to transfer funds in and out for each distribution, so getting all three distributions with just a single transfer of funds in and out was the plan here.

However, there are people who already have large amounts of USD on Bitfinex for good reason: Some people hold USD on the exchange to lend them to margin traders. Right now, I see over $16 million available for margin borrowing, and there are more currently on loan. The market lending rate is a little under 0.2%/day. (This is about 100% APY! But no free lunch here: US residents are having trouble wiring in USD because of banking regulations, and these high rates are temporary.)

These lenders can use their USD to withdraw BTC for a few hours during upcoming distribution snapshots, each of which I estimate would give them a net profit between 0.1% to 0.5% beyond what they can otherwise earn by lending them out for 2 days.

In addition, these lenders are already shouldering the risk from holding funds on Bitfinex, so I don't think they would add any significant risk from executing the plan above.

This kind of opportunity hasn't appeared before because Bitcoin's previous dominance curtailed the value of any altcoin distribution.

The mispricing here allowing for low-risk profit is in the interest rate for borrowing BTC on margin. If USD margin lenders fully take advantage of the opportunity here, then the BTC and USD interest rates should rise sharply before each distribution snapshot until there is no longer profit to be made from borrowing BTC. If this happens, I estimate that the peak BTC interest rate should be between 0.4% to 3.0% per day, depending on the expected return of the distribution round. This would be beyond historical rates on Bitfinex.

I don't expect to see any sign of this happening because this opportunity is profitable only to the few Bitfinex USD margin lenders who have a lot of USD on the exchange, and even if they execute this, I doubt their combined worth is enough to noticeably push up the BTC and USD interest rates. But we can still try to detect this by looking for large trading volumes on BTC/USD that fail to move the price, a few hours before each distribution snapshot, and an hour or so after. I don't see obvious signs of this in the historical data around the recent Byteball distribution.

I would appreciate corrections!

Profile

Stats