To stop loss or not

[Aquarians]

My observation with options is that:
1) If you sell an ATM option and try to "stop loss" through buying a further out of the money option, essentially you give up all the profit for protection. So no point stopping the loss when there's no win.
2) Only effective method to stop loss is position size. You're at peace with how much of your capital you throw on a trade and the fact that you may lose them all. You're only losing a trade, not all your money afterall.

 

[sef88]

My observation with options is that:
1) If you sell an ATM option and try to "stop loss" through buying a further out of the money option, essentially you give up all the profit for protection. So no point stopping the loss when there's no win.
2) Only effective method to stop loss is position size. You're at peace with how much of your capital you throw on a trade and the fact that you may lose them all. You're only losing a trade, not all your money afterall.

Using COVID, 2008 crisis, 1987 scenario as the worst case scenario will be a good start.

 

[Bomp]

If you are keen in a perspective from Euan Sinclair, author of several well known options trading books. He has a PhD in Physics.

https://blog.quantinsti.com/hidden-truths-stop-loss-trading/

Read the article and have the following comments:

1) Article fails to consider that when you stop, you free up capital that can be assigned to other trades. Perhaps re-enter the same trade at a better level. Both scenarios would also have an effect on the final distribution.
2) I agree that a in a long term, more or less diversifies investment, there are no need for stops. Not sure about individual stocks. Had you invested in Citi stock in 2008 you would still have lost massively(or pets.com in the dotcom bust.). Not so sure it applies to shorter term swing or day trading.
3) I use stops because it frees up my mental capital to search for other profitable trades. When all you do is try to manage a losing positions, you are not maximizing winners and finding new opportunities.
4) Related to points 1 and 3; you cant have stops of 15-20%. Those magnitudes defeat the purpose of stops.

 

[d08]

unless you are perfect in your market read-or you trade options- you need a stop.

Not really true. I use stops nowadays but that's because my positions are large. If your position is say 4% or less of your total net liq, even a 100% loss won't be catastrophic. Granted it can go over 100% with shorts but how often?

 

[sef88]

For non-options related strategy, you may consider continuous trading i.e. drop stop loss completely. You scale in and out of the positions according to your forecasts. It's an epiphany that I gotten from Robert Carver's 2 trading books (Systematic trading and Leveraged Trading). In particular, Leveraged Trading Chapter 9.

Something that I'm also trying to rationalize on my own. A paragraph that I wrote some time ago.

In conventional discrete trading, trade entries and exits are usually binary in nature, and current position size is dependent on entry and exit conditions defined t periods ago.

Current position size ~ Entry, exit conditions (dependent on current state) defined t periods ago

But in the financial world, asset returns outcome are continuous in nature with a distribution. Hence it would be optimal for current position size to be directly proportionate to Expected Returns conditional on current forecast, risk capital allocation, current volatility of instrument, overall portfolio volatility, correlation matrix, cost of rebalancing. This resonates with the Bayesian school of thought in which probability of hypothesis should be updated as more evidence becomes available.

Current position size ~ E(Returns | current forecast, risk capital, current volatility of instrument, overall portfolio volatility, correlation matrix, cost of rebalancing)

The advantage of continuous forecasts approach is that you only need to compare optimal position size given current conditions against current positions. If it diverges by x%, then you rebalance. The risk management layer and position sizing is inherently built into the framework. And it is not dependent on the state of current position. This is different from a binary trading system which is state dependent.

 

[qlai]

You scale in and out of the positions according to your forecasts.

Are you talking from perspective of overall portfolio? If you are always in position and just increase/decrease weighting, then I understand but that sounds more of a buy and hold (or managing existing position).

But when you want to be in cash most of the time and enter position only when you have a signal(swing trading for example), I think you need to make a choice of how large your initial position must be. Too small and you waste a signal that may not repeat for long time. Too large and you have no wiggle room. So it becomes a binary system.

Maybe this is a good distinction between trading and investing - do I want to be in this position or should I be in cash (aka market timing)?

 

[tradeking007yahoo]

once again stops and trailing stops etc are another way to manage your trades and your strategies. Everyones trades are different so the use of stops should be different.

the article had some good ideas but also leaned on the idea of end of day exits and or using a trailing stop to exit or losing on the trade.

Looks to me like trailing stops are not a good exit strategy unless you known your system so well that you can deploy multiple trails with different trigger points per lot.

Stops...since they are typically mkt orders they move the markets and pay the most slippage. Essentially a mkt order with a trigger price is all.

Good post here though.

 

[sef88]

Are you talking from perspective of overall portfolio? If you are always in position and just increase/decrease weighting, then I understand but that sounds more of a buy and hold (or managing existing position).

But when you want to be in cash most of the time and enter position only when you have a signal(swing trading for example), I think you need to make a choice of how large your initial position must be. Too small and you waste a signal that may not repeat for long time. Too large and you have no wiggle room. So it becomes a binary system.

Maybe this is a good distinction between trading and investing - do I want to be in this position or should I be in cash (aka market timing)?

It can be for trading or investing. You size your positions first based on volatility based measures such as standard deviation or ATRs; then you scale it up or down based on your forecasts. Let's say your forecast is only 50%, you place half a position. Few days later, when more data is in and the trend gets stronger and your model is now 80% forecast. You can scale your position up to 80%. Advantage of this is that your buying and selling of positions is not dependent on whether you are in the market or not. If you are familiar with Bayesian Statistics, this is more aligned with that school of thought.

In Robert Carver's book, for one of his forecasting techniques, he uses the gap between fast and slowing moving exponential moving averages with further normalization techniques for same unit measure across time and instruments.

You certainly can also incorporate other forecasting techniques such as breakouts, economic forecasts, ML techniques to augment your trading model.

 

[sef88]

Are you talking from perspective of overall portfolio? If you are always in position and just increase/decrease weighting, then I understand but that sounds more of a buy and hold (or managing existing position).

But when you want to be in cash most of the time and enter position only when you have a signal(swing trading for example), I think you need to make a choice of how large your initial position must be. Too small and you waste a signal that may not repeat for long time. Too large and you have no wiggle room. So it becomes a binary system.

Maybe this is a good distinction between trading and investing - do I want to be in this position or should I be in cash (aka market timing)?

Are you talking from perspective of overall portfolio? If you are always in position and just increase/decrease weighting, then I understand but that sounds more of a buy and hold (or managing existing position).

But when you want to be in cash most of the time and enter position only when you have a signal(swing trading for example), I think you need to make a choice of how large your initial position must be. Too small and you waste a signal that may not repeat for long time. Too large and you have no wiggle room. So it becomes a binary system.

Maybe this is a good distinction between trading and investing - do I want to be in this position or should I be in cash (aka market timing)?

@globalarbtrader Rob, wrote a really great post on dynamic trend following,

https://qoppac.blogspot.com/2020/12/dynamic-trend-following.html

Extracting the relevant para here,

1. We calculate an optimal position that we want to take
2. We compare it to the position we currently have
3. We adjust to get to our optimal position by trading

 

[smallfil]

Have you seen filthy rich scientists and so called experts trading and getting rich in the stockmarket? I have not either. If these BS artists believed their studies, they should trade it and make millions and become filthy rich. Why don't they? Then, they can crow about it. Ignorance is bliss. Stop losses protect your capital, profits, allow you to enter just at the right time, most times. That said, it is not a stand alone system. A trading system should have risk management which includes, position sizing. That alone will cut your risk, coupled with the stop loss. The only thing you have control over is your risk in the stockmarket. You do not know how much you will earn ahead of time. Cut your losses and let your profits run. Of course, the so called experts have no clue on how to do it!