From Cabot Growth Investor
What exactly does “building a bottom” mean? Our studies and experience tell us that the much talked about “V” bottom—that is, straight down, then straight back up—rarely happens. Sure, there’s an exact low or high to every market move, but they’re almost impossible to identify in real-time. What you can identify is the time you can safely begin to invest aggressively…it’s usually a few weeks after the ultimate low. (The only V bottom we know of happened in the days after September 11, 2001…certainly an unusually fearful time.) The vast number of great buying opportunities have come after a solid foundation was molded. In other words, bottoms are a process, not an event.
We recently looked back over a variety of market bottoms during the past 40 years and, while there is no strict rulebook, the bottom-building process generally lasts between five and 10 weeks. There are exceptions—the major bottom of the last bear market, which saw the S&P 500 fall 50%, took nearly eight months to build—but most take a couple of months. In general, the longer and deeper the bear market, the longer the bottom-building process.
During the process, the major indexes hit a low, rally sharply, and then often meander for a short period. The indexes re-test their low, often revisiting their worst levels at least once. When this happens, it’s important to watch the number of new lows on the NYSE (our Two-Second Indicator)—the readings should dry up meaningfully. At the same time, the number of good-looking stocks should drastically increase; the best will be close to new-high ground even as the market is re-testing its low.
The brief 1998 bear market provides a textbook example. The Dow slid 20% from mid-July through August 31, when it suffered a mini-crash. On that day, there were a huge 1,190 new lows. The Dow then rallied during the next three weeks, but then plunged back to its low on October 8…27 trading days (just over five trading weeks) later. The re-test brought “only” 998 new lows, and the market was off to the races afterward.
It’s the same process during longer-term bear markets. Take the monster 1973-1974 decline, when the Dow fell from 47%, peak to trough. The market bottomed on October 4, staged a nice rally for a few days, and then meandered into the middle of November. The re-test eventually came on December 9…45 trading days (nine weeks) later. As for new lows, they totaled a maximum of 661 in September of that year, but could only expand to 300 or so in December, a positive divergence.
There is no strict rulebook as to how the market will form its bottom. But the past can provide a rough roadmap to the future.