To avoid bias, hold no coins

Over the weekend, my wife and I talked money. We evaluated our current assets, shared our goals, and made a plan to invest for the future. Mainly, we set up our accounts to automatically invest in target-date retirement funds.

For the average American, “set and forget” automation is the best investment strategy. Dollar-cost-averaging into a basket of equities and bonds have historically beaten active management (e.g. hedge funds). And unless you are particularly gifted, trading your own book will yield poor results over the long run.

As we set up our system, I remembered something that Kyle Samani said in our podcast interview–that an important part of crypto investing is battling “status quo bias.” Someone suffering from this bias prefers the current state of affairs over others. Your brain thinks there is a cost associated with any change to the status quo, irrespective of the net impact of that change.

For example, if my portfolio generates a 10% return every year, I may resist improving my portfolio to one that generates 11% simply because I want things to stay the same.

Whereas status quo bias works in our favor in a “set and forget” investing plan, it works against us when we are actively managing our book–which almost every digital asset investor does.

This led me down a rabbit hole of related cognitive biases. I’ll walk through some of them here and close with some practical implications for crypto.

All sources are wikipedia unless otherwise noted.

Status quo bias

“The current baseline (or status quo) is taken as a reference point, and any change from that baseline is perceived as a loss”

Status quo bias is intuitive. We don’t like change. The act of changing or responding to change is itself a cost. We must exert energy. So it’s not entirely irrational to prefer the way things are.

But the degree to which we prefer the defaults can lead to very bad outcomes. Academic studies show that transition costs and uncertain outcomes do not fully explain status quo bias.

A paper by Samuelson and Zeckhauser called “Status Quo Bias in Decision Making” (1988) shows that:

  1. People irrationally adhere to status quo choices more than one would expect

  2. The more alternative choices available, the more pronounced the preference for the status quo choice

  3. And these biases cannot be fully explained by transition costs and uncertain outcomes

They use an example of an election to illustrate the effect:

consider an election contest between two candidates who would be expected to divide the vote evenly if neither were an incumbent (the neutral setting). Now suppose that one of these candidates is the incumbent office holder, a status generally acknowledged as a significant advantage in an election. An extrapolation of our experimental results indicates that the incumbent office holder (the status quo alternative) would claim an election victory by a margin of 59% to 41%. […] With multiple candidates in a plurality election, the status quo advantage is more dramatic. Consider a race among four candidates, each of whom would win 25% of the vote in the neutral setting. Here, the incumbent earns 38.5% of the vote, and each challenger 20.5%.

Makes you wonder about all those democratic candidates running in 2020 doesn’t it? But this isn’t a politics blog, it’s a crypto blog. So wonder about smart contract protocols instead!

If this piques your interest, I recommend reading the “applications” section of the paper. The authors walk through some examples that are relevant even today:

  • People tend to keep periodic payments (this is why the subscription business model is so good)

  • People search for new things less frequently than they should, whether it’s jobs, new restaurants, or friendships (I wrote about the related explore-exploit algorithm here)

  • Free return policies (almost nobody returns)

  • Feeling of entitlement to keep current prices, wages, or other policies leading to slower real adjustments to these markets

  • Brand loyalty

  • and more

We will come back to these later in this post.

Next up, two cognitive biases that are often used to explain status quo bias: loss aversion and endowment effect.

Loss aversion and endowment effect

people’s tendency to prefer avoiding losses to acquiring equivalent gains: it is better to not lose $5 than to find $5

Because we are just lizards trying to survive, we are more concerned about starving than thriving. Our brains are programmed to care a lot if we lose something. Just like status quo bias, this preference makes some rational sense: the greatest risk is of complete ruin (e.g. loss of life) and so anything in that direction deserves more attention than marginal gains.

But our preference for loss avoidance over gains leads to poor decision-making in situations that are not life and death.

This chart displays the relationship between value (or satisfaction) and losses/ gains.

Loss aversion

Satisfaction decreases relatively more on a five dollar loss than it increases on a five dollar gain

Loss aversion

A general rule of thumb is that the pain of a loss is twice as powerful a the pleasure of a gain.

This preference leads to the “endowment effect” where a person values a good they own more than an identical good they do not own. Their fear of losing the good they have prevents them from selling it in exchange for something of identical value. In investing, this is known as “divestiture aversion” where an investor resists selling, even if it is to buy something of equal value.

What’s interesting here is that loss aversion can play two seemingly contradictory roles in a market. The same force drives people to (1) marry their bags so to speak, and (2) sell at a loss when the market goes down.

Loss avoidance helps explain our preference for the status quo, but there’s more. “Choice-supportive bias” suggests that we actually distort reality to make the status quo seem better than it really is.

Choice-supportive bias or post-purchase rationalization

the tendency to retroactively ascribe positive attributes to an option one has selected and/or to demote the forgone options

Perhaps to protect our own psyche, we tend to come up with reasons to support the decisions we’ve already made and reasons to dislike the decisions we decided not to make.

A 2000 study from Mather and Johnson found that we remember (and sometimes construct) positive properties of chosen options and the negative properties of rejected options.

Choice-supportive bias feeds status quo bias by distorting reality to over-support the choices you’ve made that now comprise the status quo.

Practical implications

Altogether, given a status quo, we tend to overvalue what we have and undervalue potentially positive changes. This stasis is driven both by fear of loss and a tendency to distort reality to rationalize the current state of affairs.

Applied to personal investing, status quo bias leads to bagholding. Sometimes this works out (you forget about some btc on a usb stick in 2011), but other times you end up riding an 80%+ correction or watching a coin go to virtually zero. The question to ask yourself is: “if I was forced to sell everything today, would I rebuy it?”

Choice-supportive bias also helps explain the tribalism across the ecosystem. People that hold a digital asset naturally start to come up with reasons that digital asset is good and the ones they did not acquire are not so good. Person A that holds Coin A distorts reality in a way that’s inconsistent with Person B that holds Coin B. They both overemphasize their coins merits and the other coins flaws.

Bigger picture, these biases yield interesting questions about distribution.

For a digital asset to survive and grow, it needs true believers. These biases suggest that earning a spot for their digital asset in the status quo increases the likelihood that it will stay in the status quo. And the longer it stays in the status quo, the more opportunities to strengthen the belief of that holder.

Importantly, simply giving away crypto isn’t enough to earn a spot in the status quo. Anybody with an Ethereum address of a certain age has dozens of random coins that were airdropped to their wallets without their permission. But if the recipient of the coin is sufficiently aware of the coin, then maybe there’s a chance.

For example, Coinbase Earn campaigns require a user to spend some time learning about a digital asset before they receive their airdrop. Data from Nic Carter shows that at least some of these users continue to hold them after earning them.

And as many have observed (myself included) the recent trend of “initial exchange offerings” are more like airdrops than public listings given the large number of participants and relatively small allocations.

It remains to be seen what the long term effects of these airdrops are, but this line of thinking suggests that they may be more efficacious that most think (that airdrops are just sold for cash or your favorite token).

Anecdotally, I’m much more aware of XLM than I otherwise would be because I received an airdrop back in 2013/2014 and see them aidropping stuff all over the place (Coinbase, Pioneer).’s new platform Voice is best understood as an attempt to give away tokens to make EOS status quo for more people (more on this to members on Thursday).

Perhaps the winning combo for a digital asset is to (1) inspire a mass movement around a coin and (2) make it part of people’s status quo asset allocation by giving it away in exchange for a little effort from the recipient. If you can do both, you build a strong moat and feedback loop. A moat because our brains are wired to protect the status quo. And a feedback loop because many will continuously come up with reasons to support the status quo and try to get others to adopt their own status quo.

One thing is clear, hodling is a behavior reinforced by cognitive biases. This makes the most viral cryptocurrencies more viral. And if it’s not clear, it reinforces my belief that BTC or something like it is inevitable.

(Of course, there’s the irony that the thinking I’ve shared here is biased by my own support for BTC but pull that thread too far and nothing means anything)