Under the Hood #12 · A line that cannot break is a decoration Good morning everyone! Time for this week's edition of "Under the Hood." Let's get into it. Every Sunday I set a level the S&P 500 has to close above on Friday. This week I moved it down, from 7,650 to 7,500. A rule I run every week made that call, and it is the most important rule behind this account's record. What you see Most market calls cannot lose. Risks are rising is a call. So is watch for volatility. Neither one can be graded on Friday. My three weekly lines can. Each one is a number and a direction, written down on Sunday and graded at Friday's close. The record stands at 69 graded, 62 held, 7 broke. The machinery A line only means something if it can break. Put the S&P line 10 percent under the index and it holds almost every week, and the hold teaches you nothing. So I set the equity line about 2 percent under where the index closed on Friday. When the market runs away from the line, I raise it. When the line ends up right under the index, I lower it. I announce each move in that Sunday's edition, after the old line has been graded. The other two lines work differently. The credit line is a fixed trigger at 3.00 on the high yield spread, and it stays put by design. The rates line gets three quarters of last week's actual move in the 2 year yield, so a quiet week earns a tight line. The live number My file holds every weekly S&P close since 2005, 1,136 weeks. The index fell 2.1 percent or more in about one week in eight. It fell 0.12 percent or more in about four weeks in ten. Last week's line, 7,650, ended the week 0.12 percent under the index. Left there, ordinary noise would break it about four weeks in ten. At 7,500 it breaks only on a real move, about one week in eight. Since early August, with the line close to 2 percent under the index in most weeks, the equity line has broken once in ten weeks. The file says to expect about one in eight. Ten weeks is a small sample, and that match is still the first thing I check. How it breaks There are two ways this goes wrong. Lowering a line after a down week can look like moving the goalposts. My defense is that the old line gets graded first, the new distance follows the same rule every week, and both appear in the same Sunday edition. A line that held is also not a call that was right. Last week the equity line held by 9 points. That tells you the line survived the week. It does not tell you my defensive read was correct. Where it connects Credit is the trigger and positioning is the flag, and the lines are how I keep myself honest about both. This week's three are 3.00 on the high yield spread, 4.772 on the 2 year yield, and 7,500 on the S&P. All three grade at Friday's close.