Zero sum games

[Gavin Andresen]

It seems I touched a nerve when I said that getting better at playing a zero-sum game isn't a good way to make the world a better place.

I was thinking of day-traders at the time; people who don't care what asset they're buying and selling but just try to buy low and sell high to make a profit. Or high-frequency traders who try to be just a tiny bit faster executing transactions to take advantage of tiny inefficiencies in markets.

Probably I don't have a deep enough understanding of the value of liquidity in asset markets or a deep enough appreciation for their role in creating market prices, but it seems to me the world would get along just fine without them spending all their time and effort competing against each other.

I wasn't thinking of competition in general as being a zero-sum game, because in general it is not.  Competition drives efficiency, and efficiency (creating more by using less) is what makes the world a better place -- assuming that the thing you're making more efficient has a positive effect on the world.  Competing to build the most efficient bomb or assault rifle is not a world-improving activity.

To bring it back to Bitcoin: competition between Bitcoin miners is a zero-sum game for the miners, but if you think that Bitcoin will make the world a better place (I do) then the competition to be more efficient at mining is a net positive for the world as a whole.

[1714952003]

1714952003

[1714952003]

1714952003

[1714952003]

1714952003

[1714952003]

1714952003

[the joint]

It seems I touched a nerve when I said that getting better at playing a zero-sum game isn't a good way to make the world a better place.

I was thinking of day-traders at the time; people who don't care what asset they're buying and selling but just try to buy low and sell high to make a profit. Or high-frequency traders who try to be just a tiny bit faster executing transactions to take advantage of tiny inefficiencies in markets.

Probably I don't have a deep enough understanding of the value of liquidity in asset markets or a deep enough appreciation for their role in creating market prices, but it seems to me the world would get along just fine without them spending all their time and effort competing against each other.

I wasn't thinking of competition in general as being a zero-sum game, because in general it is not.  Competition drives efficiency, and efficiency (creating more by using less) is what makes the world a better place -- assuming that the thing you're making more efficient has a positive effect on the world.  Competing to build the most efficient bomb or assault rifle is not a world-improving activity.

To bring it back to Bitcoin: competition between Bitcoin miners is a zero-sum game for the miners, but if you think that Bitcoin will make the world a better place (I do) then the competition to be more efficient at mining is a net positive for the world as a whole.

Nice post.  Actually, I question whether competition drives efficiency long-term.  To me, it seems that competition for profit is primary and that one means of obtaining this objective is to provide efficient products and services.  However, long-term, I think this actually hurts efficiency.  While a given product (e.g. I-Phone/I-Pad) may be more efficient than its competitors, competition also means that this product will be outdated in a very short period of time.  How efficient is a product or service if it is already behind the curve 3-6 months after its release?  While an I-phone/I-pad may be an efficient product relative to its competitors, the process by which the product was created seems extremely inefficient as a huge amount of resources is devoted to creating a product which will become outdated very shortly.

I need to do some thinking about how this applies to Bitcoin as it is still in its infancy.

[Ichthyo]

It seems I touched a nerve when I said that getting better at playing a zero-sum game isn't a good way to make the world a better place.

Your post and point of view reminds us that we should indeed not forget the big picture.

Several of us joined the Bitcoin adventure because we realise that something is not all right with our current monetary systems, regulations and markets; so while the general idea seems ok, we think we could do better, and thus Bitcoin is worth supporting.


The whole phenomenon of trading is a by product, and I think it is a neutral one. If the Idea of Bitcoin flourishes, trading might help to provide liquidity and stabilise prices. If it doesn -- becouse it remains an isolated system without connection to the real world -- then trading in Bitcoin is indeed a zero sum game.


If we just engage into our own isolated world of Bitcoin, and forget why we engaged into that endeavour, then we might end up just creating a carbon copy of the existing economy, with all its downsides and nothing for the better.

--Ichthyo

[finway]

Anyway, Bitcoinica builds a "interest rate signal" for the Bitcoin world , at least , that's something new.

[MatthewLM]

Speculators can provide moth stability and liquidity into a market. There is an economic role for them.

[Steve]

If you sell $100,000 worth of bitcoin in the market right now, you'll push the price lower by at least $0.10.  That is not good for commerce.  So, I personally would like to see as much trading activity in Bitcoin as possible.  If leveraged trading brings in more speculators then I'm all for it.  I would love to see trading volumes increase 100x and exchange commissions get cut to 1/10th of what they currently are.

[MatthewLM]

The problem is governments and media blaming speculators for the financial crisis. People believe the propaganda. Speculation per se was never the problem.

[finway]

The problem is governments and media blaming speculators for the financial crisis. People believe the propaganda. Speculation per se was never the problem.

+1
Manipulation is the problem, like the FED.

[Steve]

The problem is governments and media blaming speculators for the financial crisis. People believe the propaganda. Speculation per se was never the problem.

+1
Manipulation is the problem, like the FED.

It's drives me nuts every time I hear some politician blaming speculators for the prices of this or that.

[benjamindees]

Competing to create the most efficient bomb gave us nuclear energy, the safest and cheapest power source mankind has ever harnessed.

Regardless, your point is correct.  Here's Tim Geithner making the Treasury more "efficient" by debasing US coins:

http://blogs.wsj.com/economics/2012/03/28/treasury-to-cut-costs-by-remaking-coins-replacing-paper/

There is only one universal measure of virtue.  It's not "efficiency," but it's not too far off.

[MarketNeutral]

It's a long shot and not all the dots are connected, but I see a potential hazard down the road if bitcoin ever prematurely attracts uninformed hedgies, quants, or any ex-prop desk traders.

Algos love zero-sum environments, even if the zero-sum environment is part of a larger, non-zero-sum environment.

The eventual quantity of bitcoins is known and fixed.

Are there other currencies and currency derivatives of which this is true?

No.

This aspect is unique to bitcoin.

This presents another dimension to the zero-sum miner dilemma.

Whether an algorithm is hedging with simple options neutrality (utilizing strangle or straddle trades, for instance) or utilizing more sophisticated trading techniques, trading using algo bots in bitcoin's environment would necessitate a reworking of nearly every algorithmic strategy I can think of. Why? Bitcoin is the only currency the eventual quantity of which is known. Quants have never had to consider anything like this within the confines of currencies. Unfortunately, I honestly can't foresee every quant reprogramming his algos to properly account for bitcoin's finite and known quantity when the allure of immediate arbitrage in a zero-sum mining environment with an already functioning algo is so tempting–––or if he simply doesn't understand the fundamentals of bitcoin's architecture and merely transplants his algo into a custom pool. For example, trading the output of an open-door pool back into the same pool is done regularly in finance (think of slush fund as a vague approximation) and, with a little modification, would work just as well with a bitcoin mining pool. Were a traditionally-programmed trading-turned-mining algo to come online, volume and liquidity could increase while masking the potential kamikaze endgame of this bitcoin-incompatible algorithm. For example, if an algo isn't designed to recognize asymptotic data, then it will only handle such data as well as it's programmed to. Does not compute. Divide by zero. Send all coins to genesis block. Open only out-of-the-money positions with infinitesimally small straddles. Reverse stops and carry decimal place. Strobe the bid and ask prices. etc. etc. Who knows how it'll respond when the future fixed quantity or finiteness inevitably becomes relevant to the trade. God forbid they're fundo algos that trade off of analyzing changes in human language. The unique degree of Bitcoin's decimal precision is another factor never considered by traditional algos. The people on the mining and technical side of bitcoin may scoff at these possibilities, because they already understand so well how the bitcoin network functions. But for finance professionals, a hash means nothing to us, because all we see is the market. Like moths to flame, heedless of just how different bitcoin is.

[Ichthyo]

The problem is governments and media blaming speculators for the financial crisis. People believe the propaganda. Speculation per se was never the problem.

Manipulation is the problem, like the FED.

I tend to disagree. Rather the problem is loosing the link to the original intention, the big picture.
Originally, all those regulatory mechanisms where created to protect society and bind the mere egoism of profit back to a larger goal.

This original reason is long forgotten, or became a shallow far fetched and remote goal. Finally we think regulation is just a value pre se, it is just good because it is good (and because we allways did it that way). Which allows requlation to be ursurped by shady, secondary and short-term goals like winning the next election.

--Ichthyo

[Andrew Vorobyov]

The problem is governments and media blaming speculators for the financial crisis. People believe the propaganda. Speculation per se was never the problem.

+1
Manipulation is the problem, like the FED.

It's drives me nuts every time I hear some politician blaming speculators for the prices of this or that.

+1

PS....  Gavin hates me

[allten]

Bitcoinica was a great service and day traders and speculators also add value to the economy.
I think where we all agree is that currently the bitcoin economy is extremely out of balance with
the majority wanting to day trade. That environment pits people against each other and there will
always be loosers.

Time and continued maturity of Bitcoin will be the only cure.

[bitcoinBull]

Probably I don't have a deep enough understanding of the value of liquidity in asset markets or a deep enough appreciation for their role in creating market prices, but it seems to me the world would get along just fine without them spending all their time and effort competing against each other.

If you accept that there is a net benefit to more competition in mining, then you should be willing to accept that there is a net benefit to more competition in trading. (as you mentioned, mining is also a zero-sum game).

More competition does not necessarily protect against formation of cartels (some recent research shows that market cartels can arise naturally without collusion). But in any case, an inefficient market might "get along just fine" but the ones hurt the most by large spreads and volatility are the non-traders (consumers, merchants and investors). As I understand it, the more inefficient the market, the more the sharks can prey on the weak.

A relatively short and highly recommended overview is The Winners and Losers of the Zero-Sum Game: The Origins of Trading Profits, Price Efficiency, and Market Liquidity.

1.2.4 Trading is a positive zero-sum game
  Rational Traders will not play a true zero-sum game in which they only value trading profits. If all traders were all alike, all expected returns would be zero and no one would benefit from trading. If some traders are more skilled than others, the skilled traders would want to trade but the unskilled traders would not. No one would trade.
  To explain why rational traders trade, we must recognize that some people trade for more than just expected profits. People trade to hedge risk, to move funds from one point in time to another, to exchange assets, to earn an unconditional expected return, to learn whether they can expect to profit from trading and to take pleasure from gambling. These external benefits make trading a positive-sum game. If the external benefits to trading are great enough, traders will trade even though they expect to lose. Skilled traders will profit to the extent that unskilled traders are willing to trade for external reasons.

[LightRider]

Competing to create the most efficient bomb gave us nuclear energy, the safest and cheapest power source mankind has ever harnessed.

Until a single nuclear plant is hit by a tsunami and radiates half the planet.

[Foxpup]

Competing to create the most efficient bomb gave us nuclear energy, the safest and cheapest power source mankind has ever harnessed.

Until a single nuclear plant is hit by a tsunami and radiates half the planet.

I think you mean irradiate (bombard with radiation), not radiate (project in all directions). You'd need more than a single nuclear plant to actually radiate half the planet.

[SpontaneousDisorder]

Competing to create the most efficient bomb gave us nuclear energy, the safest and cheapest power source mankind has ever harnessed.

Until a single nuclear plant is hit by a tsunami and radiates half the planet.

Coal kills many many thousands per year, Fukashima killed no-one.

[FreeMoney]

If you sell $100,000 worth of bitcoin in the market right now, you'll push the price lower by at least $0.10.  That is not good for commerce.  So, I personally would like to see as much trading activity in Bitcoin as possible.  If leveraged trading brings in more speculators then I'm all for it.  I would love to see trading volumes increase 100x and exchange commissions get cut to 1/10th of what they currently are.

Absolutely.

It doesn't matter if the traders care about what they are trading their offers can only help us.

[amencon]

It seems I touched a nerve when I said that getting better at playing a zero-sum game isn't a good way to make the world a better place.

I was thinking of day-traders at the time; people who don't care what asset they're buying and selling but just try to buy low and sell high to make a profit. Or high-frequency traders who try to be just a tiny bit faster executing transactions to take advantage of tiny inefficiencies in markets.

Probably I don't have a deep enough understanding of the value of liquidity in asset markets or a deep enough appreciation for their role in creating market prices, but it seems to me the world would get along just fine without them spending all their time and effort competing against each other.

I wasn't thinking of competition in general as being a zero-sum game, because in general it is not.  Competition drives efficiency, and efficiency (creating more by using less) is what makes the world a better place -- assuming that the thing you're making more efficient has a positive effect on the world.  Competing to build the most efficient bomb or assault rifle is not a world-improving activity.

To bring it back to Bitcoin: competition between Bitcoin miners is a zero-sum game for the miners, but if you think that Bitcoin will make the world a better place (I do) then the competition to be more efficient at mining is a net positive for the world as a whole.

Nice post.  Actually, I question whether competition drives efficiency long-term.  To me, it seems that competition for profit is primary and that one means of obtaining this objective is to provide efficient products and services.  However, long-term, I think this actually hurts efficiency.  While a given product (e.g. I-Phone/I-Pad) may be more efficient than its competitors, competition also means that this product will be outdated in a very short period of time.  How efficient is a product or service if it is already behind the curve 3-6 months after its release?  While an I-phone/I-pad may be an efficient product relative to its competitors, the process by which the product was created seems extremely inefficient as a huge amount of resources is devoted to creating a product which will become outdated very shortly.

I need to do some thinking about how this applies to Bitcoin as it is still in its infancy.

This doesn't sound right to me unless we are defining efficiency differently.

When that iPad is behind the curve 3-6 months after release this usually means that it will now likely be cheaper (same power for less money) or that another product with more power can be bought for a price similar to what that iPad was originally sold for (more power for same money).  How would the same iPad with no successor be more efficient in your example?  That iPad is the same whether or not another product puts it behind the curve, so I don't see how the lack of competition would change it's efficiency positively.

We could all be using 386s right now due to less or no competition and not have had to upgrade for a long time.  Through competition we efficiently produce ever more efficient products, faster and more cheaply than we would otherwise.  The faster technology improves, the faster this new technology can be leveraged to produce ever more powerful technology through a process of accelerating returns (link: http://en.wikipedia.org/wiki/Accelerating_change).

Powerful technology seems to me to be the quintessential example of efficiency since it literally allows us to do more with less effort.

In terms of bitcoin I think it is very possible that the improved efficiency to many processes directly tied to bitcoin as well as those from other technologies derived from the bitcoin concept will greatly outweigh the apparent gross inefficiencies of the mining process.  Just my opinion though, as you said it's still in it's infancy.