# Loss Aversion or Risk Aversion? from Sushil Kedia

May 28, 2013 |

Aversion to losses or aversion to risk? Which of the two is addressed by willingness and ability to close out losing trades?

Well, without invoking mathematics where it is not necessary, it is common and logical to place on the table that when a losing trade is closed one has the willingness and aversion to the risk of the persistence of loss becoming into a bigger one and one does not have aversion to the present level of loss in being accepted.

Now on the other hand, unwillingness to stop out a losing trade is indeed loss aversion.

The computations that show that having utilized some sort of mechanical rules for stopping out adverse incursions actually increased the probability of meeting with adverse incursions is totally flawed abuse of statistics.

Several arguments:

1) Historical data analysis does not undertake the "uncertainty at a given moment to decide upon" into account and is definitely incorporating hindsight 20:20 vision mind-set.

2) Any measurements of uncertainty and thus risk are never definite, since measurement of uncertainty too will be having an uncertainty of its own. So a trader in the middle of a losing trade has to decide that the level of uncertainty in his method, mind or cognition regarding the calculation of the "value of uncertainty" in his trade has become too high for him to handle. That's where humility, the currency that prevents others from profiting more from your mistake, can come into play and allow the willingness to hit the stop.

3) However, when either with or without the illusions of statistical computations of stop losses increasing the probability of meeting with more losing trades, one fails to control the human weakness of loss aversion, to somehow and anyhow turn that loss into a profit, one is becoming totally risk-insensitive. From skill, the turf changes to the power of prayer. The game begins to change from action to hope. Inconsistency of thoughts thus turns one into a trader who is continuing to hold on to risk without a mental apparatus to assess it or react to it. As the loss continues to grow not only the lack of willingness to take it hurts, the ability to accept the increasingly bigger loss also dwindles rapidly.

I am ready to be thrown before any firing squads of mathematical minds and ideas on this list if they can with or without numbers help me learn how come this list celebrates and cherishes a human value of humility and yet indulges in an idea that staying on in a trade that has incurred a level of loss greater than anticipated when the trade was opened are mutually consistent.

I would close my submission for now with one thought:

When loss aversion creeps in it makes a decision system (mind) risk-insensitive and with no respect for risk, returns are impossible. Yet, if a mind continues to be risk-averse it does not have loss-insensitivity and in humility such a mind closes out risk that has turned out to be less than comprehensible.

## Phil McDonnell responds:

Since I am the well known culprit I shall give Mr. Kedia a reply. If the probability of a decline art the end of a period of time equal to your stop is p then the probability of losing the stop amount with a stop loss strategy is 2 * p. It is simply a derived relationship. It is what it is.

It is not a misuse of statistics but rather a description of how a stop loss exit strategy will change the distribution of returns. Larry Connors studied over 200,000 trades from a winning system and compared the results with and without stops. He found the use of stops increased the probability of loss and reduced the expected gain.

In my opinion the best way to trade is to reduce position size so that no one loss hurts your account too badly. That means many small positions to me.

## Larry Williams adds:

Ahhh here I go off on a rant; please excuse a tired old mans bitterness at system vendors who claim stops hurt performance.

Yes, they are correct in that the statistics of your system will look better if one) you don't use a stop and two) your use a market with a perpetual upward bias like the stock indexes have been, usually.

They are absolutely totally incorrect in terms of living the life of a trader. So what if I am long in a position that eventually shows a profit but because I did not have a stop loss that one trade moved against be 20,000 or \$30,000 and it took a year or so to get out of? Yeah, the numbers look good (high accuracy) with no stops but it's one hell of a lifestyle.

High accuracy is a false God.

Consistency and never being in a place where you can get killed is more critical. Perhaps Mr. Connors has never sat through the reality of a large loss, especially in a large position. I have; I would rather battle the devil at midnight on a new moon with both hands tied behind my back.

It's one thing to have a system with "good numbers" it is quite another thing to be a trader and have to deal with reality.

It only takes one bullet in the chamber to kill you when playing Russian roulette. As near as I can tell trading without any stops, in any way whatsoever, is just the American version of this form of spinning the wheel.

Play the game as you wish but please heed the warnings of an old man.

## Leo Jia adds:

I have been studying the use of stops. Due to loss aversion I guess, I would like to use narrow stops. But among the various strategies I have yet found one working well with narrow stops. Good stops have to be relatively wide in my cases, but having no stops or stops that are too wide clearly hurts results (my trades are time limited). So a good choice for me is to size the position according to the stop size.

## Sushil Kedia writes:

If you reduce position size can it be argued that a position of Size N reduces to N-n implies that you took a stop loss on n lots out of N you held. Then too, it validates the fact that you do take stops.

## Anatoly Veltman writes:

Larry covered main bases (different markets, different position sizes, different lifestyles) pretty well. I just want to be sure that reader doesn't end up with wrong impression. I think the best conclusion is "it depends".

And because my act follows Larry's (who is certainly biased in favor of stops), let me try this. If you enter based on value (which is certainly against trend), then there is no justification available for a stop. Unless you argue that this stop proves you were an idiot on the entry. But if you are an idiot on value entries, then why play value…

## Anton Johnson writes:

The problem with using Conners' simulation as evidence that placing a trade stop-loss reduces returns is that he tested a winning system that likely had never experienced any 5-sigma negative excursions prior to the test date. And of course there are no guarantees that his strategy, or any unbounded trading strategy, will perpetually avoid massive drawdowns.

When implementing a strategic trade, a good compromise between profit maximization and loss mitigation can be achieved by balancing trade size along with a stop-loss, which when placed at a level that only an extreme event will trigger, will likely contain losses to a predetermined range, and also prevent getting stopped-out of a potential winner. If one is disciplined, maintaining a mental stop-loss level is preferable to an order pre-placed in the book, and available for all the bots to scan.

## Larry Williams adds:

But speaking of stops, I go back to my litany, my preaching the essential reason for never putting stops on an exchange server, or even your brokers server. Putting stops on servers means that your stop becomes part of the market. And not in a positive sort of way either. Pick a price, hit the button, and take the hit. Discipline is key here.

## Ed Stewart writes:

A trader needs a decision process for managing the expectation or expected value of the trade as well as the equity position. The problems occur when these two things are in conflict.

The thing with stops is that at times it makes no sense to get out of a trade when the expected value is still good. What is the difference between exiting at a small stop-loss point 4X in a row vs. one loss of that same size? Well, if at each "stop out" point the expected value was favorable, it makes no sense, one is just locking in losses. At times the best "next trade" is simply staying in the current trade.

However, I see Larry's point and it is a good one. Yet, the example of letting a loss get huge or holding an underwater position for a year is to me something of a false alternative. No exit strategy but hoping for a profit at some point is not a reasonable alternative.

What maters, I think, is the expected value of the trade at each moment, and balancing that against equity and a margin or error to ensure, "staying in the game".

Given this I always trade with mental stops, if not on individual positions, on total account equity. Having that "self-preservation" discipline is useful.

## Jeff Watson writes:

I learned very early on in the pit on how to go for the stops, and that weaned me off of stops completely (except in my head).

Name

Email

Website

Speak your mind