Cryptocurrency Portfolio Calculator

Should I invest in Cryptocurrencies?

The answer is - it depends. Namely, it depends on what you think is going to happen and how likely that is.

An example:

If you think it will go up by factor 10 with 20% probability and down to 0 with 80% probability, then you should invest 11% of the money you have.

How did I arrive at that number? Using the Kelly Criterion. The general formula is

share to invest = (p * f - 1) / (f - 1)

where
• f is the factor you expect your bet to multiply
• p is the probability of the multiplication

So the above example becomes:

(0.2 * 10 - 1) / 9 = (2-1)/9 = 0.111.. = 11.1%

Please note that this is NOT concrete investment advice, it's just a way to calculate how you should behave given your outlook on the future. I personally own a few cryptocurrencies, so I stand to benefit from your investment.

So how can I come up with solid values for f and p?

Well, while I don't have any idea, I do know that we have kind of a historical precedent. And I am not talking about the Dutch Tulip Bubble in the 1630s but about the Dot-com bubble.

Why?

Because we basically have a new technology (Blockchain instead of Internet) that is hyped everywhere yet few people actually use it.

So assuming this analogy is correct, we can expect a bubble top somehow similar to the one back then. One way to measure the similarity is to compare the market cap of NASDAQ (between 6.6 trillion \$ and 6.71 trillion \$) with the market cap of all cryptocurrencies available (507 billion \$ as I speak, December 14 in 2017, 16:20 UTC). You can check the current market cap on coinmarketcap.com where it is handily displayed in the top part of the landing page.

Using these numbers will lead us to believe that the whole cryptocurrency space is going to rise by factor 13.01 (= 6.6 trn / 507 bn).

Alright, so what's the value for p?

Well, that's where I don't have any idea other than I think it's possible with p >= 0.1.

But I can still come up with something useful: Assuming factor 13 is indeed going to materialize, the probability should be at the very least 1/13 = 6.6% so that the Kelly Criteration would suggest at all to invest.

Some Closing Thoughts

First of all, you should never invest more than you can afford to lose. Even if you're super-bullish, you shouldn't assign p > 50% to this. And even then and assuming a super-bullish factor of 50, your investment should not exceed 24/49 ~ 50% (assuming p=50%; f=50)

You will also need an exit strategy: One can be calculated using (again) the Kelly Criterion. You just need to calculate how much money you have (including your newly gained crypto-riches) and then calculate the percentage you should invest. If it's lower than what you actually have invested, you should reduce your position.

Another exit strategy is to just say: Look, this looks kind of right but your calculations might be off by an order of magnitude. In that case you could just define predefine thresholds where you sell part of your holdings. An approach I like would be to sell half your stake between 0.5 NASDAQ Market-Cap and 2 NASDAQ Market Cap in 20 equal steps. So you sell 2.5% at 3.3 trillion, another 2.5% at 3.8 trillion, ...until market cap reaches 13.2 (= 6.6*2 trillion) where you sold 50% of your stack (if you were lucky enough). Then reevaluate what you will do with the other 50%.

Next, I think the factor is potentially higher than the factor 13 calculated above because

• 6.6 trillion in 2000 are worth 9.45 trillion in 2018 according to this Inflation Calculator
• Government inflation figures usually underestimate the real inflation
• Interest rates are lower today than back then, at least on the global level.
• The amount of money in circulation (as can be estimated using via M2 or Central Bank Balance Sheets [PDF]) has multiplied a lot

Also, please note there is all information about the market cap included in the portfolio page (e.g., here, it also displays the factor. You can find the information at the bottom in grayish font.

And finally, just to make the point clear: THE ABOVE ANALYSIS MIGHT JUST BE PLAIN WRONG! Don't blame me if you lose lots of money ;-)

Cheers, Daniel