A PM Recalls How Steve Cohen Traded The Bursting Of The Tech Bubble

A PM Recalls How Steve Cohen Traded The Bursting Of The Tech Bubble

Submitted by DataTrek Research founder Nicholas Colas, former PM at SAC Capital

Today we have a mashup of behavioral finance and anecdotes about trading stocks at SAC during the dot com bubble implosion. While there are many similarities between frothy tech stock markets now and back then, let’s remember that the 2000 – 2002 bear market was just as much about 9-11 and the run up to the Iraq War than the end of a speculative bubble. The common lesson then to now: manage risk like we are in the early stages of a Tech stock selloff. Until proven otherwise, we are.

Our usual Story Time format typically centers on behavioral finance or market history but today we’ll combine the two and wrap them into some thoughts about current market psychology.

Let’s start with the behavioral side and discuss “Attribution Substitution”. That happens when we are faced with a complex question and, rather than do the heavy analytical work to solve it, use simple shortcuts (heuristics) based on attributes of seemingly similar situations. Behavioral psychologists like Daniel Kahneman have been looking at this topic since the 1970s, and it is a bedrock idea in the field of behavioral finance.

An example: in the last week I’ve seen two Tesla Model 3 NYC yellow cabs and one Uber on 57th Street in Manhattan. My immediate thought was “Tesla is a raging short” because whenever a car company goes into fleet sales you know organic demand is waning. That’s my heuristic from 30 years of experience covering the autos. I am substituting the “fleet sales are bad” attribution to the Tesla investment case in place of real analysis like asking taxi medallion owners if the company is offering them discounts or going to the company for an explanation.

Shifting gears to markets, a really sharp DataTrek client recently emailed in an observation that dovetails with our theme today. Paraphrasing his thought: “why do we think of the Pandemic Recession as the start of a ‘new’ cycle rather than a glitch in an ongoing upcycle driven by ever-lower rates?”

One answer – maybe THE answer – is that investors simply use heuristics like GDP growth/NBER dates, the dollar, and monetary/fiscal policy intervention as the markers for recession. A growth shock plus Fed and Federal stimulus means you reset the economic clock to midnight and begin counting out a new period of expansion. A few nights ago we mentioned the old portfolio manager sine wave model of cyclical investing where you buy Financials at the start of a cycle and Tech/Industrials at the end. That’s the melding of economic cycle “theory” and investing “practice”, but it is still based on substituting easy to find attributes for deeper analysis.

An alternative narrative, and one very much in sync with recent market action, is that the “cycle” began not on March 23rd, 2020 with the market’s lows but in October 2020 when 10-year yields finally began to rise. That makes February 2021 distinctly “early cycle” because markets have to consider what happens as rates rise. How far will they go? How much inflation are fiscal and monetary policy going to create? Will, as we’ve been highlighting repeatedly, the Fed have to increase policy rates this year? Or will they need to “do a 1994” and take rates up by 50 basis points a meeting in 2022? Unless you were (like I, Nick) in the business in the 1990s you’ve never seen a 50 bp meeting. Trust me… They are not fun.

Right or wrong, at least this narrative respects the idea that we can’t just count months and years from a market low or Fed policy shift as representative of a “cycle”.

* * *

Moving on to the sort of anecdote we often use, let’s talk about 2000 – 2002 because the bursting of the dot com bubble is another one of those heuristics people grab on to when facing complex investment decisions.

I was at SAC Capital at the time working directly for Steve so my recollections of that highly stressful period in market history are as fresh as if they happened last year.

A few thoughts from that experience that I think are especially relevant today:

#1. Unwinding the Internet 1.0 bubble took a long, long time. It did not “burst”. It leaked air, month after month after month, for years. There were actually many days in 2000 when it looked like everything would be OK. But in your heart, you knew something had broken. The old saying that “you don’t need analysts because in a bull market they’re unnecessary and in a bear market they’ll kill you” was a common refrain in the room. Wall Street endlessly reiterated their Buy recommendations on speculative tech names, to no avail.

Takeaway: remember that the S&P was only down 9 percent in 2000, and while the NASDAQ was off by 39 percent you were still up 15 percent from the start of 1999. There were plenty of opportunities to lighten up on tech names in 2000, and this was when I first heard the phrases “sell when you can, not when you have to” and “in a bear market the best sale is the first one”.

#2: It wasn’t just a loss of investor confidence in tech valuations that made 2000 – 2002 the worst bear market since 1973 – 1974. The September 11th, 2001 terror attacks hit US consumer confidence, as did higher oil prices in the run-up to the 2003 invasion of Iraq. Those dampened interest in funding cash-burning tech companies more than the first leg of the NASDAQ downdraft.

Takeaway: take away the other negative market catalysts in 2001 – 2002 and the Internet 1.0 bubble might not have burst so spectacularly. We’re thinking a lot about that just now. While speculative tech names are clearly in for some pain, they may not vaporize the way the prior generation did in the early 2000s.

#3: There’s plenty of money to be made in volatile bear markets, but you need to manage risk very aggressively. I sat next to an awesome trader in 2001 and every day was a master class in risk management. Anything that wasn’t working came off the pad quickly. Anything that was working got more capital. He wouldn’t add long exposure without an equally compelling short. When he scaled in and out of positions, he would keep his net exposure long or short exactly the same literally hour-by-hour. The stress was intense. I would often hear him muttering his mantra “I live the life I choose” repeatedly into the close. It wasn’t pretty, but it worked.

Takeaway: there are enough imbalances in US equity markets which need sorting out that 2000 – 2002 is still a good heuristic for risk management even if we don’t get a really bad bear market. As we’ve been saying, a 10 percent correction is a reasonable base case right now given the 2010 Playbook. But our heuristic toolbox includes 1994’s rate cycle, and even if it’s just a shortcut to answer a complex problem the analogy is sound enough to bring it to your attention.

Summing up with a final thought: while behavioral finance may portray heuristics like Attribution Substitution in a negative light, a big part of investing is understanding where capital will flow as a result of these very human mental shortcuts. All the chatter about bubbles (take our survey if you haven’t) and now suddenly higher long-term rates are pushing capital to the “old school company/cyclical recovery” heuristic. The first bit of that comes from what happened in 2000 – 2002. Non-tech companies outperformed. The second part is just the classic recovery heuristic.

As Steve was fond of telling us in 2000 – 2002, “don’t make things harder than they have to be.”

Tyler Durden
Sun, 02/28/2021 – 16:45

Go to Source
Author: Tyler Durden

Comments