On risk.
Risk is the possibility of an event occurring that will have an impact on the achievement of objectives. Risk is measured in terms of impact and likelihood. (Source: Wikipedia)
So much of our life is about judging and taking risks and yet we are mostly ill equipped by our educations and culture to do so effectively. Kahneman and Tversky of course were pioneers in articulating and explaining many of the basic human errors and biases and anyone who is not familiar with their work and yet interested in the subject I would strongly recommend further reading. (For a specific take on behavioral psychology with respect to financial markets, I would recommend reading the books and/or research of James Montier, an ex-colleague of mine.)
You may have noticed that I haven’t posted anything in the last few weeks. At the end of February and at the start of March, I was very very busy with work; more recently - in the last week - I have been ski racing. Unlike last year, I haven’t had nearly as much preparation but nonetheless felt confident given the training I had done that I could build on last year’s good results, especially in the speed events. And in the Downhill, I felt I could finish in the top 3. This was my key objective for the week.
The week started with a Super-G race (slightly slower than Downhill, with a few more turns) and I finished 11th, 2 seconds behind the winner. I am less comfortable with Super-G and made a couple of smallish mistakes that probably cost me a Top 5 finish, so this was a solid result on which to build. Next came the Downhill - on the same hill as the Super-G - and in the training run, I finished 8th, just under 2 seconds back. I felt extremely confident as I had not given 100% and was certain that I could improve by between 2 to 3 seconds. The winning training run time was 1:15:30. I figured that it would probably take 1:14:00 to win the race, and under 1:15:00 to be in the top 3. We headed up for the race run. The weather changed, with clouds coming in reducing the visibility. I became more confident of my predictions, and my ability to ski these times. My start number was 26, so the top ranked competitors had all finished before my turn came: 1st - 1:14:00, 2nd 1:15:47, 3rd 1:15:64… my confidence soared. The race was there to be won, and a spot on the podium was easily within reach. I went for it.
I have probably skied this Downhill course (or parts of it) almost 100 times in training and races over the last decade. I know it very well. I have never skied the first 35 seconds faster in my life. When you are racing Downhill, one of the best ways to gauge your speed is how far you fly when you go over a bump. Over the first two bumps, I went further than ever before. Over the third (and last) bump it was the same. To the extent that my usual line over the bump was now wrong: I landed too low. I needed to correct this and quickly put in my edge. But by going too low and so far, I landed in a compression (a small dip in the terrain, that has the tendency to throw you on the back of your skis.) And the light was flat. So I didn’t see it (the dip.)
As I landed in the compression and set my edge, I was thrown onto the back of my skis, caught an edge and spun out. From 90kph to zero in a dozen meters or so. I just remember two things. First, “damn - there goes the race”, and (probably a couple tenths of a second later) “don’t crash, pull it out.”
Four small elements: land too low, set edge too quickly, hit compression, flat light. Remove any one of these and I’m 90% sure nothing happens and I’m in the finish line with my best ever result 40 seconds later. As it was, I made the split second decision to not quit, start pushing with my poles and finish the race. It doesn’t take a genius to figure out that stopping in a Downhill is disastrous. I skied the rest of the course indifferently and finished 7 seconds down. This only rubbed salt in the wound - only 7 seconds down meant that I was almost certainly on track for a time below 1:16…
I was gutted.
I took a risk. It didn’t pay off. I kept asking myself over and over, “what if…?” I replayed the fateful moment a hundred times or more in my head. I didn’t sleep well that night. I couldn’t stop thinking about how I had wasted a golden opportunity.
As the week progressed, I skied ‘ok’ but never really had what I thought was a really good run. My results were more or less in line with my results from last year. I was disappointed. But as the week came to a close, my son helped me see things in a different light. He asked me how I thought I had done and if I was happy with how the week had gone. Obvious questions, but answering them made me change my perspective. Overall, I was clearly disappointed not to have done better, not to have improved on last year’s results; and in particular not to have nailed my run on the Downhill. But on the other hand, I had not skied particularly well and yet was on a par with my results last year when I thought I had skied well. My base level had clearly improved.
And I was forced to ask myself: “What if I hadn’t skied so aggressively in the Downhill? Not made the mistake and finished 5th? or 10th? Would I have been just as disappointed? How would I have felt knowing that had I attacked the course, I might have won?
I realized that I had made the right decision. The only thing to do was to attack. To take risks. To seize the opportunity. The upside of winning for me was much higher that the downside of not finishing in the top 10. Even though I lost. Especially because I lost.
Taking risk means losing sometimes. The market’s estimation of the probability of a given loss is implicit in option, betting or insurance pricing. But a probability is just that - any individual outcome is either 0 or 1. Even if the probability of 0 is 5%, on one event the outcome can obviously be 0. Which doesn’t mean the probability was wrong (or right) ex-post. Or that probabilities are of no use in risk management.
It means that one needs to understand the relative consequences of a bad and good outcome in absolute terms ahead of applying some more or less robust estimation of the probability of each. I think this is where a lot of traders and financial organizations go wrong. They look at a bad outcome - say losing $1bn - and estimate the probability of this occurring - say 5% - and come up with a probability adjusted loss of $50mn. They look at a good outcome - say making $100mn - and estimate the probability of this occurring - say 95% - and come up with a probability adjusted gain for the trade of $45mn. And then they ask, can we afford to lose $50mn on this trade? (Of course I am over-simplifying, but not to the extent that it alters the point…)
Of course from time to time, the bad outcome will happen (despite having only a probability of 5%) and the trade will lose $1bn. There is no such thing as a realized probability-adjusted loss or gain…
When deciding whether or not to take a particular risk, first you need to understand how the worst case outcome will affect you. Next you need to understand if the upside potential is worth the downside risk. Finally, when you lose - which is inevitable - even with 99% sure things, you need to have no regrets. Put it behind you and move on. ‘Back-trading’ yesterday’s risks is the surest way to ensure failure.
And so it isn’t hard to jump to the conclusion that the best measure of a trader, is her attitude and actions in the face of losses. The true measure of one’s character is taken in the face of adversity. It is not an original idea but worth remembering from time to time. Especially in times like these.
