From The Advisor’s Fuel Podcast with Adam Koós, CFP®, CMT®, CEPA — a conversation with David Keller, CMT, founder of Sierra Alpha Research and host of the Market Misbehavior podcast.
Most advisors were taught that paying attention to what the market is actually doing is a bad idea.
Sell the narrative. Build the thesis. Treat price as an afterthought that either validates your story or doesn’t. David Keller spent nearly two decades unwinding that training, first as director of research at Fidelity, then teaching graduate students, and now mentoring advisors through Sierra Alpha Research. His argument is simple: price carries information the same way earnings do, markets trend because humans are predictably irrational, and the advisors who win aren’t the ones with the best predictions. They’re the ones with the best routines. Here’s what came out of the conversation.
Because price has value, and ignoring it is how good analysis still loses money.
Keller’s favorite way to make the point is a question he asks every room: have you ever found a company with a great management team, a solid growth story, strong earnings expectations, bought the stock, and watched it go down? Every hand goes up. What got missed wasn’t the fundamentals. It was the fact that price itself is information.
He frames it through the P/E ratio, a lesson from technical analyst Craig Johnson. Fundamental analysis is the work of defining the E: projecting how a company will grow earnings over some forward period. Price is the other half of the ratio, and it carries its own signal. Technical analysis is simply reading that signal, analyzing historical price action to make informed assumptions about what may come next.
And this isn’t fringe anymore. The growth of momentum investing has put academic rigor behind what technicians understood for decades: markets tend to trend, and those trends tend to persist. Search the momentum factor and you’ll find peer-reviewed studies backing it up. If a finance professor taught you markets are perfectly efficient, Keller’s response is blunt: they aren’t. He used to play his graduate students audio from the trading pit during a flash crash and ask whether the sound of pure panic sounded like rational actors calmly exchanging assets at fair value.
That it’s a crystal ball. It isn’t, and treating it like one is where investors go wrong.
Keller’s reframe is that technical analysis is less a way to predict the future and more a microscope and a satellite: a tool for seeing clearly what’s happening right now and what’s working versus what isn’t. The value isn’t in forecasting tomorrow. It’s in acknowledging what’s actually in front of you, because most bad decisions come from holding onto something far too long and refusing to admit the picture has changed.
It’s a distinction Adam’s firm lives by too, describing themselves as trend followers, not trend predictors. Or, as Keller puts it, quoting a line often attributed to Tom Dorsey: there’s nothing wrong with being wrong. Just don’t stay wrong.
The two that do the most damage are confirmation bias and the endowment effect, and advisors face a double dose of both.
Client conversations, Keller notes, are a petri dish of behavioral biases: people want to buy too early, buy too late, sell too early, sell too late, all driven by emotion. But an advisor actively managing portfolios has their own biases to manage on top of that. Two stand out:
The antidote to both is the same thing: a consistent, evidence-based routine for evaluating market conditions, so the process, not the emotion, drives the decision.
A change of character is when the market’s look and feel shifts, and catching it is most of the game.
Keller organizes his work around spotting these shifts. The market carries a certain rhythm and texture during a strong uptrend, and then it starts behaving differently: divergences show up in breadth, in momentum, in sentiment. Much of reading these is lagging and trend-following by nature, which is the point. You’re not trying to call the top in advance. You’re recognizing, based on evidence, when the trend that was working has stopped working.
His ideal is to combine leading indicators (things like seasonality that help you anticipate what could happen next) with a healthy dose of lagging indicators that confirm a cycle has genuinely turned. Both matter. Neither is a prediction.
Because clients want two things, and only one of them is a financial plan.
Adam’s framing: clients want to know they’ll be okay when things go sideways, and they want to know their advisor has their back and is doing something about it. When markets convulse and a client calls asking what the plan is, “stay the course, see you Monday” is a hard answer to give with a straight face. A rules-based technical process gives the advisor a real answer, and gives the client evidence that someone is paying attention.
The risk-management version of this is one of Adam’s favorite analogies: the signals that a market might crash are like tornado sirens, and the crash itself is the tornado. Not every siren produces a tornado. But when one goes off, you probably still want to head to the basement.
You accept that being wrong is part of the job, and you make sure you don’t stay wrong.
Keller is refreshingly direct about this. Even with a great process, the market sometimes won’t do what you expect. He recalls how painful an early lesson this was for Fidelity analysts who had gotten straight A’s at elite schools and suddenly found themselves in an industry where they’d be wrong often. Peter Lynch, he notes, was wrong plenty; the difference was that when Lynch was right, he stayed right, and when he was wrong, he didn’t stay wrong for long. Lynch’s gardening image captures it: water your flowers and pull your weeds. Nurture what’s working; cut what isn’t. The losing move is the reverse, pulling your flowers and watering your weeds, selling winners and doubling down on losers, which is how a portfolio dies a quick death.
There’s nothing wrong with a whipsaw taking you out and putting you back in, either. Getting back in at higher prices isn’t a failure of the process; staying in a losing position when the evidence says otherwise is. As a friend once told Adam: the day you accept you’ll always buy or sell a little too early, too late, too much, or too little is the day you’ve got the whole thing figured out. A strategy that wins 60 to 70% of the time and loses only 30 to 40% is a genuinely excellent one.
Routines. Full stop. It comes down to how you spend your time.
This was Keller’s clearest single answer of the conversation. What do you do every day, every week, every month, every quarter to make sense of the markets, evaluate performance, and make changes when needed? His guiding principle: a consistent but imperfect process beats an inconsistent perfect one. Don’t chase perfection first; build consistency, then improve incrementally.
In practice, that looks like:
Keller borrows a pilot’s term for this (he’s a pilot himself): the pre-flight checklist. Adam adopted the same concept, crediting Keller and advisor David Cox. A pre-flight checklist for the day, week, or month forces you to see what’s actually going on before you act.
The deeper point: behavioral biases create problems precisely when the right routines aren’t in place. Bull markets are dangerous here, because you can get away with bad habits when everything is going up. Everyone’s a genius in a bull market. That’s exactly why the bull market is the time to upgrade your routines, because by the time the market turns lower, it’s far too late to start building good habits.
Three tools do most of the work: relative strength, momentum, and moving averages.
Asked what habit the next generation of advisors should build, Keller’s answer was humility, the kind that comes from treating markets as a lifelong learning opportunity.
The most successful investors he’s worked with, from Steve Cohen-level hedge fund managers down to people trading a few thousand dollars, weren’t the smartest people in the room. They were the most self-aware: crystal clear on what they’re good at, honest about what they’re not, and surrounded by tools and resources to cover the gaps. He tells the story of sitting in company meetings with portfolio managers who’d beaten their benchmarks for decades, expecting them to have all the answers, and finding instead that they asked the best questions. The best investors, it turned out, weren’t the ones who knew everything. They were the ones still learning.
Keller’s whole message, that a consistent, evidence-based routine beats emotion and prediction every time, is exactly the kind of discipline that separates a practice that plateaus from one that scales. If you want to build that kind of process into your own advisory business:
For more on rules-based defense you can actually execute, see our conversation on market breadth and risk management with Vincent Randazzo. And for the behavioral side of portfolio discipline, try why probabilities beat Cinderella stories.