Mental Accounting for Crypto Investing: the Safe and the Adventure Chest

Investing in cryptocurrencies presents some new psychological problems you rarely see from investing in traditional instruments, like equities and bonds.

  1. Uncertain risk profile. It’s not clear how risky cryptocurrencies are. Whereas we think we have good historical data of how well equities and bonds will do, there isn’t agreement on how we should value cryptocurrency markets. Moreover, counterparty risk from holding cryptocurrencies is significant. There’s the chance that cryptocurrency exchanges shut down: Mt. Gox infamously declared bankruptcy, mining services like NiceHash have been hacked, and exchanges have shut down because of regulatory pressure. It’s also difficult to weigh the convenience, liquidity, and exposure of buying coins on different exchanges against holding a physical wallet or running a full node, then forgetting your keys or having it robbed.
  2. Mental accounting bias. The lack of clear valuation bounds for crypto has made as many uplifting rags-to-riches stories as sad ones. Paying constant attention to the ups and downs of cryptocurrency markets takes a serious toll, especially when traders perceive their fear of missed opportunities as a loss. For example, see this painful apocryphal story on r/Bitcoin/ of an enthusiast committing suicide after missing out on $50M.

How, then, do you effectively invest in a growing array of cryptocurrencies, altcoins, and newcoming derivative instruments, without losing your mind in the process? This is where behavioral economics can help.


Here’s a teaser:

  • Case 1: Imagine that you have decide to watch a movie tonight and you already purchased a ticket for $10. As you enter the theater, you discover that you have lost the ticket. The ticket office tells you they keep no records so you will have to purchase a new ticket to see the movie. Would you still pay $10 for another ticket?
  • Case 2: Imagine that you have decided to watch a movie tonight and the ticket is $10. As you enter the theater, you discover that you have lost a $10 bill on the way. Would you still pay $10 for a ticket for the movie?

Between both cases your answers should be equivalent: you’ve lost $10, either in the form of a $10 bill or a $10 ticket. But when psychologists Kahneman and Tversky (1984) asked survey respondents these questions, far more people were willing to buy a new ticket in the second case.

It turns out that a dollar isn’t a dollar. People tend to build different budgets in which spending is treated differently. So, while people were treating the second case as just losing a separate $10 from their general net worth, they were treating the first case as spending $20 on the movie — which feels like paying more than in the second case.

The same bias applies to investing: behavioral biases make us panic more than we rationally should, or act in ways that contradict our beliefs.

These are some case studies that I’ve encountered:

Years ago, A. bought several Bitcoins at $400, and lost it all due to an exchange scam. At $5000, he thinks that the price of Bitcoin will rise, but is reluctant to buy or mine Bitcoin because he’s already “spent too much money on cryptocurrencies.”
  • While lessons can be gleaned from a former loss is a good reason to improve your security, or use physical wallets, your current investment decisions should be independent of a previous loss in the same domain.
C. bought Ethereum at two price points: $200 and $400. Believing that the price of Ethereum might dip below $400 because of a news event, she sells only the tokens she purchased at $400 so that she “wouldn’t lose on any part of her portfolio.”
  • Tokens are fungible: a token bought at one price is the same as one bought at another. Your decision of how much to sell should be independent of the price you bought tokens at, but dependent on how certain you are of the price movement. Regardless, even experienced investors base decisions on transaction utility instead of the bottom line (Thaler 1985).
D. is convinced that the value of Bitcoin will rise over time. However, she feels that it is “too late to get in” as the price has already risen by a lot.
  • During a rally, prices may be anchored to its previous price history. The investment becomes perceived as a loss against what she could have gained in retrospect — even though the investment would still be a net gain.
F. is convinced that the value of Bitcoin will rise over time. However, the price of Bitcoin suddenly falls because an exchange was hacked. While still convinced of the long-run value of Bitcoin, he panics and sells his Bitcoin until it returns to its previous price.
  • Investors are known to overweight unexpected, short-term losses, and are known to be impatient in “locking in” prices — even when they result in trading fees or losses.

All of these failure modes are affected by mental accounting bias, among others. When we extrapolate the wrong conclusions from our mental accounts, we fail to make rational decisions on the margin. Even when these biases don’t translate into losses, they take up our attention span and direct us to the wrong questions.

Now that we know mental accounting bias exists, we can use it to our advantage. On a very natural level, we’re using mental accounts to rationalise our investment decisions and manage self-control. We can frame our investment decisions in a way that protects us from the outsized psychological impact of unexpected events and losses, so we can make rational decisions.


We introduce a two-portfolio mental accounting partition for thinking about investment decisions: the Safe and the Adventure Chest.

The Safe is an investment portfolio you design to be low-risk, stable, and to absorb the psychological impact of any unexpected losses on the Adventure Chest. The Adventure Chest is a portfolio you design to be high-risk and high-reward, for exposing yourself to more adventurous ideas that require constant attention. The holdings of each should be roughly proportional.

When you file an investment to your mental Safe, you resolve to make sure you’ll have continued access to it (passwords, authentication, insurance, etc.), then stop meddling with it. When you try something out in your Adventure Chest, you want to be able to adjust your exposure quickly.

If you lose money in your Adventure Chest, you should treat your money as disposable: you’ll have done it for your entertainment and self-learning, and it’ll be offset by gains from investments in the Safe. If you win a lot of money in your Adventure Chest, you’ll want to transfer some of that money you make into instruments you can keep in the Safe so a proportion is maintained.

Why does this work?

  1. It’s a heuristic for diversification. Thinking explicitly about two portfolios with two different purposes forces you to diversify your holdings: whether it’s investments, exchanges, or security methods. This makes you more robust to losses to any part of the crypto ecosystem.
  2. The Safe absorbs the psychological impact of the Adventure Chest. Pledging to stop caring about a partition of your investments directs your attention to items that need it the most, and reduces the chances that you’ll make panicked, irrational decisions. The impact of a catastrophic loss in the Adventure Chest would also be blunted by gains in the Safe.
  3. An explicit partition forces you to look at the big picture. By forcing yourself to decide whether an investment is low- or high-attentive, you can decide more quickly whether it deserves your risk exposure.

This approach is psychological, so it’s blind to what your investments are, or what your risk profile looks like.

Here is what the distribution of investments might look like for a newcomer to cryptocurrencies:

Safe (80%)
- Employer-matched IRA (Vanguard Target Retirement Fund)
- Bond funds
Adventure Chest (20%)
- Equity investments
- Bitcoin (Coinbase)

For a cryptocurrency enthusiast who is almost certain that Bitcoin prices will rise in the long run:

Safe (70%)
-Employer-matched IRA
- Bitcoin (hardware wallet)
Adventure Chest (30%)
- Cryptokitties
- Cryptocurrencies (Binance; Bitstamp)

Here’s what a typical response might look like:

All cryptocurrency prices crash due to the closing of an exchange.

  • The newcomer reinvests more money into Bitcoin (if he still believes in the long-run value of Bitcoin) or equity investments. The impact of this event on the enthusiast is blunted due to the gains in his IRA and bond funds.
  • Similarly, the enthusiast is less likely to panic-sell the Bitcoin in her Safe unless the crash changes her belief in the long-term value in cryptocurrencies.

Bitcoin eventually becomes an institutional standard for value storage.

  • The newcomer sells some of his Bitcoin and reinvests in his bond fund.
  • The enthusiast decides to buy a hardware wallet to store her bitcoin, and uses a password manager with 2FA to store Bitcoin on multiple exchanges. She transfers some of her Bitcoin exposure out to the Adventure Chest in order to explore decentralised applications, altcoins, and arbitrage opportunities.

Brian Ng is a data scientist and poet who has a degree in Economics from the University of Chicago. Previously he was an economics consultant at TGG Group, where he worked with economists as Steven Levitt and Daniel Kahneman on behavioral economics and data analysis consulting projects.