Suggested by ScilabGuruP

Last edited by Paul on January 10th, 2004, 11:00 pm, edited 1 time in total.

According to a very professional trader, and a highly respected contributor to this forum, there is no difference. I'll let him come forward and explain his point of view.

- adannenberg
**Posts:**284**Joined:**

In gambling the rules of the game are defined, i.e. there's no model risk. Not so in asset pricing, and hence in trading...

- ScilabGuru
**Posts:**297**Joined:**

QuoteIn gambling the rules of the game are defined, i.e. there's no model risk. Not so in asset pricing, and hence in trading... This is why gambling can be considered as a simple model of the market where all models are true models. So if you don't understand the gambling rules/risk/models - then don't think about trading too. I met so many situations when people are talking about very sofisticated models, measures, bla, bla, bla and cannot solve the simplest probability riddle from the gambling world. (Actually I am not an exeption - I like gambling puzzles but not all of them are easy for me)

- ScilabGuru
**Posts:**297**Joined:**

When everything is known from model's point of view (probabilities and payoffs) the optimal derived strategy becomes a pure mathematical problem. Coming back to the trading we split our problem into two 1) Assume that the model we have is the "true" model, or more preciesely - estimate probabily that our model is the "true" model. (In gambling this prob. is equal 1) 2) Derive a trading strategy conditionally to the above assumption. Then having set of different models one can construct an "portfolio" of different startegies weighted by the probability of the model.

- DominicConnor
**Posts:**11684**Joined:**

A gambler plays with his own money, a trader plays with his employer's.This gives rise to a variety of agency effects, which drive rational behaviour.A traders risk is bounded from below by zero, so can never leave with less wealth than he started.A gambler does not have such a hedged position, and may lose not only the money he entered with, but may be granted credit to lose all his assets, whether explicitly wagered or not.It follows that one may expect a trader to be less risk averse than a gambler.Firms combat this agency effect by imposing controls, both formal and qualitative. This directly gives rise to more complex agency conflicts.Most professional gamblers return is simply a function of the cash acquired or lost. A trader derives different returns based upon how profit or loss was made.For a trader a virtuous profit will result in a greater personal return, conversely a loss made in a way sanctioned by his employer will impact his rewards less.Since such constraints are arbitrary and based upon personal judgement, a rational trader will expend a proportion of his time in explaing the virtue of his actions in a way directly that makes them more personally rewarding. It follows that the tighter the constraints imposed by an employer, the greater the proportion of trader effort is expened on spin rather than attempts to trade profitably. A successful gambler acquires by these activities more capital to invest, acquired in part from unsuccessful gamblers. Since a casino or other house is taking the other side of a bet, it can also be modelled as a gambler, albeit one with more favourable odds. A gambler is free to invest whatever % of his capital he feels appropriate inThis funding model is very different to that experienced by traders. The nature of investment firms is that unlike gamblers, capital is diverted away from succesful traders towards less succeful ones. This is referred to by bank management as "strategy", however the consensus amongst practitioners is that this is another word for agency activities of senior bankers.The risk aversion of traders vs gamblers is not fully captured by a simple coefficient. Consider the case of a trader who is in a position where his trading has left him with a deficit.If this is beyond a certain limit, his continued ability to earn will be compromised or eliminated. Thus the rational course of action is for him to become as risk loving as he can within the constraints imposed upon him. A gambler by contrast finding that he is consistently losing may simply quit the game.It follows from this that a trader's earnings are more serially correlated, since a gambler doe not usually rely upon external funding for his investments. A trader has a certain amount of reputational capital, which may be valued as his ability to earn money relative to a new trader. Few gamblers have any such utility for reputation, and in the rare cases described in [Thorpe, 1966] external constraints often ensure that gamblers with reputational capital actually reduce their abiltiy to extract value from their market.A trader may observe opportunities to extract value that are within the rules imposed by his employer, but which are not seen as generall virtuous by his employer.He will then avoid such trades reducing the return to both him and his employer, solely to preserve reputational capital.Gamblers have no such constraint, and are neutral to the notion of any restraint other than procedural or legal. Indeed the personal utility of a gambler may be increased by the opportunity to show his prowess by doing something on the edge.A successful trader will be given more opportunities to trade, however a successful gambler will find that other participants in his market decline to gamble with him.

Let's make a market with the underlying: payout per contract 100 if there is no evidence of any difference between gambling and trading, contract term: dec expiry, contract size: 100I'll start the market: 90 at 110 (in thousands)we'll see where the market goes... and don't forget the market is always right!

- ScilabGuru
**Posts:**297**Joined:**

I did not tell that there is no any difference. Of course there is. I try to tell that the gambling is somehow model of trade where all probability structure is known perfectly. As it was told by Adannenberg there is no model risk here. This is as seems to me the most important point. The other condtions (investor preference, employer/gambler/trader relation) can be changed for gambler to make him exactly as trader.

Not sure if I agree there. In Vegas, there is also plenty of betting on horseraces, professional sports, where the "odds" are not known exactly by any mortal (except the boxer who accepted a bribe to throw the match), and even slot machines are not transparent in their odds to the gambler. Even the guy who sits at all the baseball games and takes exact hit statistics faces quite a bit of model risk, similar to those faced by the managers in Michael Lewis's Moneyball.How can we answer the difference between someone who bets on the snowfall in Central park on http://www.tradesports.com versus a trader of weather futures? It seems that the underlying and the openness of its probabilities is not enough to differentiate the two activities...

- ScilabGuru
**Posts:**297**Joined:**

Yep, I agree, that it is too strong to say that we know all structure with parameters in the gambling, but at least the number of states is finite and the structure is known (parameters are not if this is not the simplest gambling). The number of explanatory factors is also countable.

Comparably so, though. The states of the payoff of a bond are also finite (bondholders receive 100, or 99.99, or 99.98, ... or 50.26, ... each with a discrete probability).I woudn't say the factors in a horserace are necessarily more countable than the factors affecting AAA commercial paper. In both trading and gambling there are simpler and more complex cases, payoffs, probabilities, and clarity of information. I don't play slots, especially now that they have gotten so computerized and complex, but still play poker, craps, baccarat, and other games that are more transparent.To me, a more important dimension is "why"? I go to Vegas, Atlantic City, Monte Carlo, Copenhagen, or Macau for fun and entertainment, and expect to lose my "cost of entertainment". When I trade an option, I am spotting a risk that I have an advantage in bearing relative to my counterparty, and expect to make money on average.

- ScilabGuru
**Posts:**297**Joined:**

I heard a rumor that abouit a year ago Pentagon wanted to open an exchange on political events. (To find Ben Laden, Saddam, president elections, etc). It was refused by Congress(??, not sure by whom) due to ethical and probably some other reasons.

So is TradeSports a gambling or trading site?