Lots of people took that exact same strategy at what in hindsight now looks like ridiculously low prices like $1 or $10 or $100. Those people don’t have 50,000 or often any bitcoin left.

Sell 1/4 of your holding every time the price doubles. You won’t make as much money as you could ideally, but you will make a sizable fraction of the maximum, no matter how far or near it will go.

> you will make a sizable fraction of the maximum, no matter how far or near it will go

That won't be true if the price goes high enough. By realistic counter example:

So you start when BTC is at about 10c. You mine around 2000 bitcoins (not bad, $200 for using your old 5870 that was collecting dust). Your card burns out so you quit mining, but implement the above strategy.

You sell 500 at 20c, 375 at 40c, 280 at 80c, etc.

Once bitcoin hits $13k, you have about 10btc left and have made $32k out of your possible gains of $20M.

I would say that less than 1 percent is not a sizeable fraction.

Of course, you would make a sizeable fraction if the price weren't able to double so many times, but the above example is actually quite realistic for some.

Fair.

Proposed amendment: once you start fearing that bitcoin will crash and wipe out your crypto wealth, begin to implement the "2x 1/4" algorithm. This rule will prevent excessive bleeding at the tail.

I posit in your scenario I would start worrying at $50/btc ($100k), so that's 8 doublings till 13k and by that time I am still holding 10% [(3/4)^8] of my coins (worth $2.6m) plus I have 1.9M in cash [1] I took off the table for a total of $4.5m. Worst case (btc drops instead of going up) I still made $25k.

  [1] https://www.wolframalpha.com/input/?i=sum+(2000%2A(3%2F4)%5Ej%2F4%2A50%2A2%5Ej),+j+%3D+0..8

The problem is picking those points.

At about 30c, I was pretty sure bitcoin would crash, wiping out several hundred dollars which would recover the sunk cost of the GPU.

At $7 was another point I was pretty sure it would crash (and hey, getting $10k out to pay down a car seems like a heck of a lot of money for an internet peculiarity I did no work for).

After $30 when it dropped and Mt. Gox was dying, I was pretty sure it would be dead for good, as were many others.

I don't think your amendment really helps because of two facts:

1. "Fearing it will crash" is something that is difficult to gauge, and I felt since day 1 of bitcoin it was likely to crash and wipe out enough money to pay for a few months of rent or more.

2. Your assumption about that working still fails if it simply doubles more, e.g. if bitcoin goes to $1M/btc, we now have a similar "less than a percent" of possible profit, so you still don't get a significant fraction... And in fact if BTC takes off like that, it likely means the USD will have crashed and thus your previous gains are even less useful.

It's easy to look in hindsight and say "selling fractions from the 10s or 100s would have been a good strategy if we knew bitcoin would stabilize in the 1000s", but such hindsight didn't exist back then, and we still can't be sure it's fully stabilized.

I'm not saying it's a bad strategy; it does seem pretty reasonable. I'm just claiming that it won't guarantee that you end up with a significant fraction of the maximum possible gains, and in fact I don't think any such strategy exists.

When I say "upset" and "fear" I mean something significant - you ought to be upset about it enough to be still upset a year later. "Could have paid for the video card" is not sufficiently upset. "Could have bought a car" would be if you couldn't afford a car to start with. Whatever it is that makes a difference in your life is where you start.

And on the upper end I am restricting myself with "diminishing marginal utility of money" (aka FU money). The range between "too little" and "FU money" is quite narrow, logarithmically speaking.

Let's say that the lower boundary is $100k and FU money is $20m, and same starting conditions as before. 12 iterations would give me $9.7m in cash plus $13m in BTC [2000(3/4)^1250*2^12].

I admit I have deviated from my original "sizable fraction" statement, and I no longer believe in it. But having shifted the goal posts I proclaim these new ones are better - just as satisfying, but a lot more achievable.

That's basically what I did and I ended up with tens of thousands instead of millions (if I had HODLed).

Then again, I've been beset with illness and medical bills, needing to pay for multiple surgeries, so I had little choice.

I presume that the first step of your strategy would be to sell 1/4 of your holding now, since it's already doubled (more than once).

I also presume that you mean, the second time, that you sell 1/4 of the 3/4 you have left.

That strategy works for something like Bitcoin, if you got in near the beginning. It's a really nice strategy for that situation.

But let's say your employer gave you got stock options at $1/share which, at the time they were granted, was exactly the stock price. The stock slowly goes up to $2.50, then slowly falls to below $1.

There are two problems with your strategy in this scenario: First, when has it "doubled"? You were given these options for free, and (on the day they were granted), they are worth zero. As soon as the stock goes to $1.01, they've more than doubled... but you don't want to sell then. (This is also true of Bitcoins that you mined back when that was easy. They were essentially free, so when was the first doubling?)

But if you wait until the stock is worth $2 ("doubling" in another sense), then you only sell once, and only a quarter of your holdings. That's much less gain than you could have made.

Your strategy is a good one for a rocketship like Bitcoin. But you often don't know that you're on that kind of a rocketship until it's doubled several times. If you wait for confirmation that you're on a rocketship, you could miss some much smaller, but still really sweet gains.

What I did when I had the options scenario is essentially your strategy, but with much lower numbers. When the stock went up maybe 25% or 50%, I sold off some of the options. And some more when it went up another 25%. That worked very well for me. But if the stock had been a rocketship, I would have missed out on a lot.

Some of this is bound by the size of the investment. Sub 10,000$ returns are not going to change your life, but when you can start cashing in the equivalent of your annual salary not taking money off the table gets risky.

At that point think of it as a math problem.

Suppose you own a stock that’s never going to issue a dividend, but will do buybacks until eventually the company fails. In such situations you can expect 0-X doublings and a final price of zero.

Now, what strategy works best depends on X, buy and hold is clearly a losing strategy in all situations. Further you don’t know X but you can come up with a reasonable bounds as nothing is going to be worth say 10^30$.

Another approach is to simply constantly sell a fixed percentage each month. That’s very useful if you only have one large investment that does not nessisarily fail, just stop growing exponentially.

If you’re going to rebalance your BTC-cash portfolio every time the price doubles you should probably use Kelly bets, right?

Figures someone already gave it more thought than I did. Good to know!

my general thinking about such meteoric rise of, really, any security like btc/amzn/etc is that this is a cognitive mistake our mind does because of labels. the label amazon wasn't 'Amazon' until it showed consistent growth & shrewd judgement over time. the mistake we make is take that mental object we see now and back project on the past while remorse turns off our analytical ability to see all the possible risks associated with it then.

I always remind myself that most public information is already priced into a stock & for every buyer of a stock there is a seller too.

Yep, it's gambling.

I sold at $15 and was happy that I'd paid for some real fancy graphics cards with the proceeds. Valuation at one point was well over $500k, oh well. Oh and half went poof in MtGox.

I paid for a month's rent with a wallet I'd found with only a transaction fee inside which somehow didn't get deducted from a few years back.

Well lucky you mtgox is getting a civil rehabilitation

True. Had I been involved with Bitcoin, I probably would've done that, myself, and regretted it.

But then again, if we're talking about cashing out for an amount of money on which you can pursue hobbies, live comfortably, and never have to work for another person again... I can't imagine I would regret it that much.

I had bitcoin worth about $50 in 2009/10 go poof when I reimaged a laptop and forgot to backup.

And you habe not been killed by frozen poop falling from an airplane in the sky. Which would have been as likely as Bitcoin becoming a speculative hype. Celebrate that!

Absolutely. A few months earlier I lucked out by selling a ton of stock/funds when I consolidated retirement accounts (I didn’t like the trade restriction during the transfer period). I missed the crash as a result.

Back of the napkin level calculation says that if you had about 675 bitcoin in 2010, it would've been worth $10,625,000 if you'd sold it at the $17,000 peak.

Bitcoin traded between $0.003 and $0.39 in 2009-2010. Source: http://bitcoin.zorinaq.com/price/ Therefore "$50" worth back then could be any amount between 128 and 16667 BTC, worth between $800k and $103M today.

You weren’t kidding about back of the napkin because that’s not 675 * 17k.

He clearly calculated for 625 BTC (it divides evenly to the precise multi-million dollar amount) but had either a typo in his post or in the calculating.

Yes, I typoed 625.

I’m very glad you were wrong. :)

I never calculated the actual potential amount, because I figured it would be depressing, but 9-figures would be beyond depressing.

I’m not sure what I actually paid because I bought it from a friend directly. It was somewhere around 100 BTC. Definitely a cautionary tale about backups and bad luck. But on the other hand, I’ve in whole been the recipient to more than my share of good fortune.

I put the down payment of about eight grand on my last car with Bitcoin I sold in January of 2012. So you can do the math. :)

Now seems like a reasonable time... I mean they have to sell eventually, right? What are they waiting for? Trillions?

Your assets are only really worth something when you sell them. Many more should have sold at the height of the boom last year.

While true, that's hardly actionable. It's impossible to repeatedly pick peaks.

Yeah, it's like saying everyone should have shorted housing in 2006.

If everyone had perfect information... markets wouldn't even function in a recognizable way. If everyone sold before the peak there wouldn't be any peaks.

I guess it all depends on whether you think BTC is more or less "real" than dollars or these supposed "securities" of which you speak.

Maybe they already sold enough to be comfortable for decades and are just trying to make the space welcoming and fun, because if they were into Bitcoin early on they probably love the idea of it and don't just want to make money, they want people to use it.