A subscriber brought up a topic that has troubled all of us options traders during our trading careers. Should stops be set according to the price of the underlying or the option? Should some other parameter be used?

I promised to cover the topic but warned that it was a complex one that would require some time to develop. After thinking over the topic for a couple of weeks, my thoughts boil down to one question: What were you thinking when you placed the order?

Annotated Ten-Minute Chart of the SPX on 7/16:

What if, instead of taking off the hedge upon a trendline break, I'd been watching for a trendline break as a signal to enter a bearish options trade? If this had been my thinking, then the whole bearish trade would have been predicated on that trendline break.

Using a trendline, chart formation, or moving average break to initiate a bearish or bullish trade means that you've predicated this breakout trade on this basis. The trendline, chart formation or moving average now provides a logical initial stop.

For example, that trendline break occurred at about 1088.50. The initial stop for a bearish directional trade entered at that point would logically go back above the trendline. The chart doesn't have the indicator Average True Range (ATR) displayed, but at the time of the break, Average True Range (ATR) was reading at about 3.2 points over a 10-minute bar, if I'm eyeballing it correctly. I wouldn't want to be stopped on normal gyrations during a puny 10-minute period, so to get outside the noise of a 10-minute period, I might choose to set my stop slightly above 1088.50 + 3.2 = 1091.70.

With a strong drop like that one, I would understand that the ATR was likely to expand as things get more volatile, so perhaps I might add $0.25-0.30 to be sure I didn't get stopped out on normal noise for the next 10-minute period. What I would not do, however, is add $0.30 and set my stop exactly at 1092.00. Why? So many people set their stops at round-number levels. If I really thought I ought to give it that extra $0.30, then I'd give it a little more and set my stop at something like 1092.17.

Now we get to another aspect of the "What were you thinking?" question, however. What if you have a $5,000 account and chose to entered that trade with an August 1060 put, a put with a delta of 0.3564 or 35.64 if the 100 multiplier is applied? That put would have theoretically cost $22.59 or $2,259 plus commissions for the entire trade. That's not a choice I would have made with a $5,000 account, with SPY puts being a better choice for that size account, but I'm choosing an extreme example to make a point. The move back up to the theoretical stop at 1092.17 would have been a 3.67-point move. In the absence of other changes, that bounce would have theoretically erased $3.67 x 0.3564 = $1.31 from the value of that option. That would have been $131 plus commissions. The loss would have been greater if both trades had been off the mid price, buying higher than the mid and selling lower than the mid. It's my experience that when prices suddenly break through an important support or resistance level, those puts or calls violently increase in price for a few moments, so those breakout directional trades often mean that you're buying expensive options. I wouldn't be surprised if those puts had actually cost $1.00 more than that that theoretical price.

However, let's use the $131 plus commissions cost theoretical loss if the SPX moved back up to the initial hard stop. Let's figure in a commission of $1.50 per trade, which is a typical cost for a Think-or-Swim commission, making the loss at least $134 and probably more. Maybe far more. If one has a $5,000 account, a $134 loss would be a 2.68 percent loss of the total account.

Calculating whether a stop is set appropriately means also calculating whether the resultant loss would be appropriate for the account size. In this case, is that $134 or 2.68 percent loss an acceptable loss per trade? Most would consider it an acceptable loss for a single trade, with figures from 2-5 percent of the account total per trade being quoted as appropriate losses by various sources. In this case, of course, I would argue that putting that much money at risk with that expensive of an option was not an acceptable risk to take, but we're talking about how we set stops here, and I was making a point using an extreme example.

The second part of that "What were you thinking?" question must be how much of a loss are you able to sustain and still keep outside the worse risk-of-ruin scenarios? It doesn't matter where logic dictates that you put a stop if that stop level subjects your account to too big a loss.

Calculating where a stop should be placed often proves to be a two-part question involving both the movement of the underlying and the price of the option. Sometimes they can't be separated. If the most logical stop will produce a loss that's unacceptably large, such as 10 percent of the account total, then the trade shouldn't be entered at all. The stop is too far away, no matter how close it seems when viewing the underlying's movements.

When I first began trading, I read an article which suggested that if an option had lost a third of its value, the trade probably wasn't working. I used that parameter for a while, but that parameter doesn't factor in how expensive the option was in the first place and what the size of the trader's account is in the second place. How much of a dollar-amount loss you can sustain must always be part of the decision before a trade is entered. If the most logical stop would result in a loss bigger than you can sustain, the trade shouldn't be entered.

This discussion so far has centered on trades initiated on the basis of trendline, chart formation or moving-average violations. Trades initiated on this basis probably require hard stops based on the underlying's price, as long as the appropriate dollar-amount loss for one's account is heeded, too. A trader who initiated a trade on the basis of a trendline, chart formation or moving-average violation, either to the upside or downside, is thinking that the violation means that a certain type of trade is appropriate. If the underlying's price then reverses that breakout, the trader's thinking is proven wrong and the trade isn't working.

Another time when the most logical stops would be based on the price of the underlying would occur when the trader is thinking that support or resistance in the form of previous swing highs or lows will hold rather than break. For example, right before the bar in which prices broke through that trendline, prices had closely approached it. RSI had reached a level from which it often reverses. We had experienced many mornings when an initial downturn had been met with buying, with one occurring just the day before the breakdown. A trader with a bullish bias might have thought that close approach to the trendline was a good opportunity to enter a bullish trade. If so, that trader would soon have been proven wrong. This time, a stop might have been set at least 3.2 points below the location of the trendline at 1088.50, so at 1085.30 or lower. That stop would have been hit on the same 10-minute bar in which the trade was entered.

The takeaway point here is that initial stops can be based on a technical analysis setup, such as a trendline, formation or moving average. If so, the trader has the ability to easily set a logical initial stop based on that formation. Such stops should probably not be set so tightly that normal market noise--the normal jits and jots of prices as they bounce around a bit--will hit the stop, but always, always the underlying profit or loss should be kept in mind. Determining what is normal market noise has proven extremely difficult in this market environment. Traders who heed those logical stops have often have found their stops hit just before the market reverses and takes off in the direction the trader was hoping to play. Traders who don't heed their stops are sometimes punished even more severely. Imagine the case of the trader with the bullish sentiment who entered a bullish trade just before the trendline break and didn't set or heed a logical first stop. There's just no way around it: sometimes the stops you set are going to be hit just before the market reverses, no matter where you set them, but good risk management is still necessary.

I've barely touched on the subject of setting initial stops in a couple of types of trades, so more will come in succeeding articles. We haven't yet touched on what kind of stops might be logical for a trader who enters a relatively delta-neutral type of complex option trade such as a butterfly, calendar or iron condor. We haven't discussed what would happen if a trader had entered a trade based on a certain volatility setup: a trader who entered a butterfly just before an earnings announcement, anticipating a sudden decrease in volatility that would benefit the butterfly, for example. We haven't yet discussed what to do after an initial move that's produced significant profit. Where should the stop go then?

I won't always know the answer. I don't trade volatility plays much, especially earnings announcements. Although I traded a lot of pure call or put trades a decade ago, I mainly use long calls or puts to hedge risk in my complex options trades these days. I will have entirely different parameters for determining when to enter those trades. In any case, I don't present myself as the expert on these pages but instead as someone with at least a little experience who is exploring what there is to know about options and inviting you along on my journey. However, what I do know for certain is this: before you can set logical stops--whether based on the price of the underlying or the price of the option--you must remember what you were thinking when you entered the trade and have an idea of what proves to you that original thinking was flawed.