Modeling economic growth and zero-sum games in Bitcoin

lotteryI have a new article that went up over at Let’s Talk Bitcoin yesterday called, Can Bitcoin change from a bubble economy into a growth economy? (pdf).  It is essentially part 2 to the previous piece and put together based on feedback I received this past week.

I should mention that while I did use an equation in the middle of the article to describe prices, I do not think the regression itself is valid.  As my friend RD mentioned, the error term does not seem random. There is a deterministic trend that would have to be filtered out first.  It is the same problem with modeling long term GDP growth.  After all, why would we need all these economic models if we can simply draw a straight line and predict GDP using high school math?  Many things follow an exponential growth curve, that is nothing new. The exciting thing is to forecast it in the short term, which this method is extremely bad at (the same criticism can be lobbed at models like Elliot wave theory).

Over the last 12 hours I have received some criticism about one particular point: zero-sum games and gambling.

In one email exchange, Bob disagreed that gambling was zero-sum, stating:

Gambling in and of itself does not create productivity, but the businesses that surround gambling certainly can. See: casinos, bitcoin mixing services.

Also, if the house makes a bunch of money, and the owning entrepreneur uses those funds to start another, productive business, then in some sense the gambling has facilitated economic development by liberating wealth from unproductive suckers who participate in online gambling to a highly productive entrepreneur. This was, of course, exactly the case with the most popular bitcoin gambling website.

The issue here is a measurable one.  A zero-sum game is one in which wealth is merely redistributed and not grown.  What Bob described above is economic activity but not economic growth.  Gambling is zero-sum game as is speculating on stocks or cryptocoins, no new utility itself is created.  Tokens are simply being moved from person to person. Eventually many people are left with assets that they cannot sell because all the demand has been fulfilled, and at that point the price may actually crash.  In other words, to make money in a zero-sum game, it is only because others have lost an equal amount.  In fact, in many cases, value diminishes because of interchange fees or in the case of mixing services, transaction fees.

This touches on an economic principle of opportunity costs (the “seen” and “unseen”) — the traditional example used is Alice throwing a brick through a shop keepers window.  While the seen result is a repairman being hired to fix the window, thus spurring economic activity, this does not actually create economic growth because the shop keeper must now forgo certain opportunities to spend repairing existing physical stock.

Note: Gambling has the name “math tax” because it is a tax on people who are not good with statistics (49.5% odds means in the long-run, you will always lose to the house).  This is derived from Ambrose Bierce’s quote, “Lottery: A tax on people who are bad at math.”

After a quick Google scholar search, I think there is more concise explanation of this phenomenon in Gambling and speculation, by Borna & Lowry:

Unproductive nature of gambling

For players, gambling, at best, is a zero-sum game, i.e., the aggregate wealth of the players will not be altered due to a gambling activity. The losses of one party are precisely equal to the gains of the other participants. Of course, if the gambling activity were taxed by the government, or there were other ‘leakages,’ then the expected value of winning would be negative, i.e., the aggregate wealth of the players after the play would not be equal to their original wealth.

Although gambling is a sterile transfer of money or goods among individuals creating no new money or goods, it nevertheless consumes the players’ time and resources and may subtract from the national income. From a macro-economic point of view, the aggregate wealth of the players will change, in the long run, due to the fact that the transfer of wealth is usually among unequal productive sources. It may be argued that the productivity lost due to a transfer of money from one player will be offset by an increase in the productivity of the other player. This assumption is true only if both the winners’ and losers’ production schedules were assumed to be identical and linear.

One last note: there are a number of people I would like to thank for their comments included in this article; this does not mean they agree with or endorse my view.  This includes: Dave Babbitt, RD, Mark DeWeaver, Dave Hudson, JL, Taariq Lewis, CK, Petri Kajander and Chris Turlica.

2 thoughts on “Modeling economic growth and zero-sum games in Bitcoin

  1. Interesting. My question is, does zero-sum matter? After all, such calculations only take into account economics. It is zero-sum economically.

    Economics isn’t the whole story though. We live in a way more complex system than that. Couldn’t the re-distribution from suckers to people who understand complexity have indirect growth in terms of greater innovation down the line? Additionally, what about all those other factors?

    Cryptos teach people about finance, and how corrupt the stock markets are. They build communities. They inspire. They make life interesting for people who previously didn’t have a stake, not just in economics, but in *life*.

    Economics is only part of the story. In such a case, zero-sum game becomes a very limited, and not particularly useful, insight in my opinion.

  2. Here’s an example. It may be zero-sum, but it’s creating an entire class of “new-money”. The last time that happened on this potential scale was the rise of the merchant class in early modern Europe.

    It won’t be so much ‘new economic growth’ as it will be ‘new economic paradigm’. One that rejects growth as the primary indicator to strive for, imo.

Leave a Reply

Your email address will not be published. Required fields are marked *