VALUE: After Hours (S08 E06): Is Buying the Dip Still Smart?

During their recent episode, Taylor, Carlisle, and Juan Torres discussed Is Buying the Dip Still Smart? Here’s an excerpt from the episode:

[Jake] Sure thing. I thought that Juan might get a kick out of this week’s veggies, so I’ve been saving it for a little bit. I want to look at a behaviour that really feels like it’s been trained into markets over the last decade and a half, which is buying the dip.

At first, BTFD was the thesis. You’d see weakness. Maybe you’d assess the policy, balance sheets, valuations, recession risk, and you decide whether the expected payoff justified that risk.

Maybe even try to price the Fed put. What was that worth? But when a behaviour gets rewarded repeatedly, it stops being a thesis and starts becoming just a natural reflex.

This isn’t because people are necessarily lazy. It’s because our learning systems compress what works into something fast and quite literally calorically cheap to execute. Here’s the question that really matters.

What happens when the payoff function changes, but the cues stay the same? In markets, this core payoff function for buying the dip has been dominated by the price of money. Real rates, liquidity, discount rate, all that’s applied.

It’s all been going down. We had a Zerp period forever. When the price of money falls, duration gets rewarded, multiples expand, and drawdowns are shorter and shallower.

When the price of money rises and stays up, that mechanical tailwind obviously weakens or maybe even disappears. That same drawdown can be a completely different animal. There’s actually a clean laboratory analogue for all this, which was the psychologist Anthony Dickinson in the early 1980s.

He had rats where he trained them in a simple box. They press the lever and they get a little food pellet. It’s a very stable contingency of repeatable trials.

Over time, these rats learn the rule and they press very efficiently. They push the button, they get a treat. Push the button, get a treat.

But then Dickinson devalued the reward. Outside of the box, the same pellets were then paired with a nausea-inducing chemical. When the rats got sick from eating the pellet, they recovered.

Then afterward, they avoided that pellet when it was offered directly to them outside of the box. They clearly knew that the pellet had become undesirable. Well, now here’s the critical test.

They put them back in the lever box after this. If the lever pressing was goal-directed, pressing in order to obtain a specific outcome, then pressing should decline once that outcome has been devalued. We see the outcome and we say, okay, we don’t want that anymore.

Don’t push the button. For many rats, that’s exactly what happened. They reduced their pressing.

But the over-trained rats, those that had pressed hundreds of times for this magic button, they often did not. In fact, they just kept on pressing at old rate, even though they would refuse the pellet when it was handed to them outside of the box. Same knowledge, but very different control system, actually.

This is a really key distinction of what’s happening, goal-directed behaviour versus habitual behaviour. Goal-directed behaviour is outcome-sensitive. It uses a model in your head of, if I do X, Y will happen.

Then it updates that value as Y changes. In the brain, this kind of evaluation leans on your cortical, your newer part of your brain. It’s involved in planning and inference.

It helps to represent, what is this outcome worth right now? That’s your system two thinking, really. Now, habitual control is different.

It’s stimulus and then response. The cue appears, the action fires. It’s learned through repetition and reinforcement.

It can run with minimal evaluation of the current value of the income. It’s that system one thinking. The basal ganglia is part of your brain that is at the heart of this.

These things are not just for movement. They’re also selection machinery for action. With training, control can shift from more associative outcome-sensitive pathways towards more habitual pathways.

You’re just going to basically hit the bid. Let’s map this back to the markets. For a long stretch after the GFC, the environment has just repeatedly rewarded this one pattern, like buy every single weakness, catch the rebound.

Monetary policy has been accommodated this whole time. Liquidity is quite abundant. Rates trended lower, multiples expanded.

The cue basically was the drawdown. The market would have a tantrum about it. The response was buying and all the rats hitting the lever.

The Fed would panic and then intervene. Then your reward, your pellet that you got to eat was this recovery. It just kept happening over and over again.

When was the last time we didn’t see that play out? The important thing to think about, too, is this is not really stupidity or lack of discipline that’s happening. These poor rats were overtrained.

They weren’t confused. They knew that the pellet was bad. It’s just the part of the control in the brain that was driving the behaviour.

The context was different. I think that might be a useful way to understand why regime shifts can feel sudden, even when the ingredients have been changing for quite a while. These habits don’t update just because new information exists.

They update through experience, usually through repeated prediction errors. One failed rebound might get shrugged off. Two starts to get rationalised.

It takes a series of disappointments before the system reallocates control back to the deliberate evaluation. My hypothesis is that it’s going to require a fair amount of abuse to desensitise all of the dip buyers that have been pushing the little lever for the last 15 years. Here’s a quick quiz to bring Toby and Juan back into this segment and outside of my soliloquy.

Multiple choice, no pressure. If a rat refuses the pellet when it’s handed to it directly but keeps pressing the lever in the box, what’s most likely going on? Is it A, it forgets what the lever does?

B, it’s hungry and just being a drama queen? C, the behaviour has become habitual and isn’t updating to the new value of the reward? Or D, it’s trying to keep collecting its rat management fees?

[Tobias] Ah, good one, JT.

[Jake] C.

[Tobias] C for mine.

[Jake] Of course. Everybody knows the answer is C. I just like to get jokes in once in a while.

Once you see the structure of this, hopefully there’s some practical implications. The point is not never buy the dip. The point is don’t confuse a trained response with an evaluated decision.

You can find out more about the VALUE: After Hours Podcast here – VALUE: After Hours Podcast. You can also listen to the podcast on your favorite podcast platforms here:

VALUE: After Hours (S08 E06): Is Buying the Dip Still Smart?

Read next