The margin is not the limit

On a deep dive, the no-decompression limit is the number printed on the table for the depth I am planning to spend time at. At twenty-one metres on air, my body can absorb dissolved nitrogen for about forty minutes before I have to start staging decompression stops on the way up. Below forty minutes at that depth, I do not need stops. Above forty minutes, I do.

This number is a margin. It is the margin between the dive a healthy person can complete safely and the dive that, statistically, starts to produce decompression sickness in some non-trivial fraction of similar dives. The margin is computed conservatively. It is a number set in advance by people whose job was to think about exactly this problem, in conditions much worse than mine.

A new diver, looking at the table for the first time, learns two things. The first thing is the limit. The second thing, which the diver learns more quietly, is that the limit is conservative.

The conservative limit becomes the dangerous limit the moment the diver treats the margin as an opportunity.

I want to write about this, because I have watched myself, repeatedly, treat the margin as the opportunity instead of the protection. The mistake feels small at the time. It looks like sound reasoning. It looks like efficiency. It is the most common mistake I have made in diving, and I have started to notice the same shape in the way I think about positions, money, and time.

The mistake works like this. I look at the table and see that at twenty-one metres, I have forty minutes before I have to start staging stops. I plan a dive for twenty-five minutes of bottom time. The dive proceeds. At minute twenty-three, I have visibility, I have good air, my buddy is fine, and the wall opens out into a small cave I would like to look into. The cave will take five more minutes. Five more minutes at twenty-one metres brings me to thirty minutes of bottom time. Thirty is still well below forty. I am still inside the margin. I look at the cave.

The arithmetic is correct. The reasoning is wrong. The reasoning is wrong in a way that is not visible while you are doing it.

The arithmetic is correct because thirty minutes at twenty-one metres is well inside the no-deco limit. I do not need to stage stops. The dive will not produce decompression sickness in any statistically meaningful sense. The boat will be there. The ascent will be normal. Everything will be fine.

The reasoning is wrong because the limit was not designed to be used as a moving target. The limit was designed to be the line you do not cross. The plan was the line you set inside the limit. The space between the plan and the limit is the margin, and the margin is not yours to spend. The margin is the protection for the failure mode you do not yet know about.

If I were to extend my dive from twenty-five minutes to thirty minutes, and the regulator started free-flowing at minute twenty-seven, and the controlled ascent took ninety seconds longer than expected because of the panic transition, and my buddy needed two minutes of buddy breathing, the margin between thirty minutes and forty minutes is what absorbs the unplanned cost. If I have already spent the margin to look at the cave, the failure mode that I did not plan for does not have anywhere to land. It lands on me.

The margin is not yours, because the margin is the part of the dive that is reserved for the events that you do not yet know are going to happen.

This is the rule I have been working hardest to install. The plan is the limit. The table is the limit on the limit. The margin between the plan and the table is the protection for the dive you did not plan, not the room for the dive you did plan to extend.

The market version of this is the cleanest analogy I have been able to find anywhere in my writing, and I want to be specific about it.

I have a written rule that no single position will exceed twelve percent of my portfolio. The maximum is twelve. The rule is set conservatively. The reason the maximum is conservative is that twelve percent in any one position means that a fifty percent loss in that position is six percent off the whole portfolio, which is a loss I am willing to take but not willing to take often. The twelve percent is the table.

The plan is the position size I actually enter at. The plan is set before I enter. The plan, in most cases, is six or seven percent. There is daylight between the plan and the table. The daylight is the margin.

What I have to refuse to do is treat the daylight as opportunity.

The temptation, in a position that is working, is to add to it. The reasoning is identical to the cave reasoning. The position is up. The thesis still holds. The market is calm. My maximum allowed size is twelve percent. I am at seven. I have five percentage points of margin to spend. Why would I not spend them, when the trade is working.

The arithmetic is correct. The reasoning is wrong. The five percentage points of daylight are not the room I have to grow the position into. The five percentage points are the protection for the position I do not yet know is going to break. The market version of the free-flowing regulator is the gap down in the morning that I did not anticipate. The market version of the buddy needing my octopus is the position I had to hold longer than expected because the thesis took longer to play out than I budgeted. The margin between the plan and the table is the absorption for the events I did not plan. The margin is not the position size I get to grow into.

The wrong version of the rule, the version most people carry in their heads, is: my maximum allowed size is twelve percent. The right version of the rule is: my planned size is six percent, my maximum allowed size is twelve percent, and the gap exists to absorb events I did not plan. I am not allowed to enter at twelve. I am not allowed to add into twelve. The gap is not mine.

This is the rule the title of the essay is trying to install. The margin is not the limit. The margin is the protection for the part of the dive I am about to have that I did not predict.

There is a real place where this analogy fails, and I have to acknowledge it.

In diving, the limit is set by physiology. The no-decompression curve is the same for me as it is for any other diver of similar build at similar exertion. The conservative margin is computed against a population. Whether or not I personally spend my margin does not change the curve. The curve was there before I was there. It will be there after.

In markets, the limit is set by me. There is no physiological curve. I picked twelve percent because of how I want my portfolio to behave during a fifty percent drawdown of a single position. I could have picked ten. I could have picked fifteen. The limit is internal. The margin is internal. The whole thing is something I wrote.

This is the part of the analogy that has to be named. In diving, the rule is not yours. In markets, the rule is yours. The temptation to spend the margin in markets is stronger, because in markets you also have the authority to rewrite the rule. You can convince yourself the rule was conservative, then change the rule, then enter at the new limit, which is the old limit's table.

The only cure for this that I have found that survives writing is to refuse to change the rule in the direction of any current position. The rule changes are allowed. They are not allowed on a day when changing the rule would let me hold a position I could not otherwise hold. The rule changes happen when nothing is at stake. The rule changes happen on Sundays, in writing, with a record of the previous version, with a sentence explaining why the change is being made, and a commitment to wait one week between the writing and the change taking effect.

This is the buddy check on the limit. It is not the same as the buddy check on the position. The buddy check on the position asks: am I inside the rule. The buddy check on the limit asks: am I rewriting the rule to make a position legal, or am I rewriting the rule for a reason that has nothing to do with any current position.

The dive version of this is simpler, because the table is fixed. I cannot rewrite the no-deco limit on a Sunday. The limit is the limit. The plan is the part I can change.

This is the part of my analogy that makes diving the cheaper teacher and markets the more dangerous one. In diving, the table is not negotiable. The only thing I can move is my plan. In markets, both are negotiable. The margin is the thing the new version of me, who has spent the day looking at a position, is most likely to reach into.

The cure is the same as the cure for the unpunished violation. Friction. Slowness. Refusal to interpret a calm sea as evidence about the table. And, above all, the refusal to spend the margin in advance for any event other than the one the margin was reserved for.

The margin is not the limit. The margin is the part of the rule that is not yours.