When I first got interested in trading,
A big one – focusing on techniques instead of edges.
I poured time and energy into building backtesting engines and frameworks for applying all the usual tools – cointegration tests, Hurst exponents, Kalman filters, neural networks.
You know the drill.
I was lucky enough to know some people who were a few steps ahead on the journey, and one conversation literally changed the trajectory of my life.
I was proudly showing off a machine learning framework that I’d been obsessing over, confidently expecting to impress this person with my technical prowess (arrogant, much?).
Instead, they simply asked, “What’s your edge?”
My blank look prompted a reframing of the question.
“What effect are you trying to model?”
I didn’t know what to say.
In my mind, the system was the edge – although I didn’t even realise I was making that assumption at the time.
Pressing further, they asked why this specific thing would make money. Of course, I had no good answer.
The brutal truth is that I felt really clever using all these “quant” tools and techniques. But without putting edge first, I might as well have been drawing lines on a chart or conjuring Elliot Waves.
It’s like an architect spending months designing the perfect kitchen for a house built on quicksand. The details are meaningless when the core concept is fundamentally flawed.
If the foundations are solid, the right kitchen can make a meaningful difference. But on its own, that fancy kitchen isn’t saving a doomed house from collapse.
Edge comes first. Everything else is just implementation.
What Is Edge, Really?
Edge is the return you’d expect if you could take the same trade infinite times. Any single trade can go either way, but with true edge, you’ll make more than you lose over the long run.
It’s what statistics people would call “positive expected value,” or what a casino operator would call the difference between the true odds and what they pay out to the punters.
It’s an abstract concept. But it plays out in the real world:
Market returns are dominated by unpredictable randomness in the short term. Even perfect trades with perfect information will have wildly variable outcomes.
Imagine you have an edge in predicting when Tesla is going up. If you take that trade often enough, you’ll make money. But any single trade could tank if Elon goes on Joe Rogan and smokes weed.
This randomness means:
- A good trade isn’t necessarily one that makes money
- A bad trade isn’t necessarily one that loses money
- It’s only good or bad when considered in the aggregate – does it have positive expected value?
Where to Find Positive Expected Value
Markets are both incredibly simple and overwhelmingly complex at the same time.
On the one hand, it’s literally just people buying and selling stuff.
But those people have all sorts of goals, constraints, time horizons, infrastructure, and personalities. It’s competitive, adaptive, adversarial, and cooperative, all at once.
And while it’s hard to untangle causality at the micro scale, good mental models and organising principles can help make sense of things at a scale that might be of practical use.
In that spirit, here are four guiding principles for where edge arises. Of these, only the last two are practical targets for most independent traders.
1. Arbitrage Relationships
The same things should have the same prices. When they don’t, you can buy cheap, sell expensive, and capture the difference – theoretically at least.
Why we probably can’t do this: Pure arbitrage is incredibly competitive. By the time you spot these opportunities, they’re gone. The exception is brand new, immature markets where infrastructure is primitive, for example, latency arbitrage on new crypto DEXs. But even if you find somewhere you can compete, expect it to get harder over time.
2. Information Advantage
Predictable things get priced in. This is fundamental analysis – building models of companies, economic outcomes, asset valuations and forward expectations. It’s also consensus. The aggregate market is a ruthlessly efficient expectations-pricing machine.
Why we probably can’t do this: You need to know things better than very smart, well-resourced people who are also trying to know things. And not only do you need to be right, you need to be more right than the rest of the market (ruthlessly efficient expectations pricing and all that). Unless you have genuine specialised expertise, this is exceptionally difficult.
3. Risk Preferences
People dislike nasty stuff. They’ll sell you uncomfortable investments at discounts and overpay for comfortable ones.
Why this works: This isn’t about being smarter than everyone else – it’s about being willing to accept risks others won’t and managing them well. The edge persists because human risk preferences persist. Think of this as a win-win transfer of risk – not a better prediction.
4. Flow Effects
Buying and selling pushes prices. When someone needs or wants to trade large size quickly, it moves markets. This creates opportunities to buy things cheap and sell them rich.
Why this works: Large (relative to liquidity), price-insensitive trading creates distortions. Someone trading for reasons other than maximising trading returns creates opportunities for those who understand what’s happening.
Only (3) and (4) offer realistic opportunities for most of us.
Risk Premium Harvesting: A Win-Win Foundation
This is our best idea (“best” in the sense of most confident in, most likely to persist): buy nasty stuff, sell comfy stuff. Get paid for taking risks others don’t want.
Here’s how it works:
Imagine a bond with $1,000 face value. Say there’s a 95% chance of full repayment, and a 5% chance of total loss. The mathematical fair value is $950, but it will trade at $900 because people hate losing money – they won’t pay what it’s intrinsically worth.
That $50 discount is your risk premium for bearing risk others won’t.
Examples include:
- Equity risk premium: Stocks return more than bonds because they can crash harder (exposed to more risks)
- Credit risk premium: Corporate bonds yield more than government bonds because companies have higher default risk than governments (generally)
- Volatility risk premium: People overpay for downside protection (think insurance)
- Term risk premium: Longer bonds yield more (usually) because they’re more volatile (exposed to more risks over a longer period)
These strategies typically generate Sharpe ratios around 0.5, give or take. They show significant volatility relative to returns, frequent losing periods, and occasional large drawdowns.
This discomfort isn’t a bug; it’s a feature. If the returns were smooth, everyone would pursue them, and the premiums would disappear.
No risk, no premium.
For most of us, risk premia harvesting should be the foundation. You don’t need to be smarter than the market – just willing and able to hold risks others find uncomfortable.
Getting an Edge in Other People’s Trading
You might hear people talking about “flow” – this just refers to trades hitting the market.
If you can find pockets of price-insensitive flow – people willing to trade at any price – then you have a chance of getting an edge from other people making things mispriced.
The key insight is that when someone needs or wants to get something done, the market makes it expensive for them. If you really need to transact in size or quickly, you’re going to pay for it.
That’s why you hear people who’ve been around for a while say, “Why did the market go up? More buyers than sellers.”
They’re not being facetious; they’re speaking honestly to an uncomplicated truth – if you have more aggressive buyers than passive sellers willing to sell at good prices, the market is going up. There’s value in not overcomplicating this.
People trade for reasons that have nothing to do with price or value:
- Forced liquidations and margin calls
- Index rebalancing requirements
- Redemption-driven selling
- Regulatory constraints
- Positional imbalances
- FOMO
- Coming home drunk and punting copper futures (Ahem. Speaking for a friend here.)
When this forced or price-insensitive trading happens in size (relative to liquidity), it pushes price away from fair value, creating opportunities for those who understand what’s happening.
And by “understand,” I don’t mean untangling causality at the level of understanding exactly who’s doing the trading and when. What I mean is, having a plausible explanation for why this trading might be happening, why it might be so large as to cause a dislocation, and why others haven’t completely absorbed it. Preferably, those explanations would be backed up with some data.
Three Examples of Flow Trading
Flow opportunities follow the same pattern: forced trading pushes price away from where it should be, and you take advantage of the mispricing. Here are some examples.
Example 1: The Reversion Wrench
The idea: React to something being mispriced.
The approach: Observe something already mispriced and fade it back to fair value.
Example: Interest rate futures typically form smooth curves. When one contract trades way rich to the term structure, you could sell the expensive one and buy surrounding contracts, betting on reversion.
Example 2: The Market Neutraliser
The idea: Preemptively position ahead of predictable flow, then supply to it.
The approach: Suspect that forced trading is coming, position ahead of it, and provide liquidity to it when it comes in.
Example: End-of-month Treasury effects. Institutional buying typically pushes bond prices up at month-end, then they revert. Buy bonds before the month-end, sell after the turn.
Example 3: The Positioning Opportunist
The idea: Join positioning imbalances, then fade them.
The approach: Temporarily join positioning-driven moves rather than immediately fading them.
Example: Short squeezes. When heavily shorted assets start rising, short sellers face margin calls and other forces to cover, pushing prices higher. Rather than fading immediately, you ride the forced covering, then become a liquidity provider.
Why You and I Can Compete (sometimes)
Unlike arbitrage (too competitive) or competing on information (too hard), risk premia harvesting and flow trading approaches are realistic for indie traders.
We can make money here because:
For risk premium harvesting:
- Competition is low or non-existent – you’re not actually competing, you’re transferring risk
- It works at any scale
- It’s based on human psychology, not temporary inefficiencies
For flow trading:
- Some opportunities are too small for larger traders
- Some require operational flexibility to move quickly
- Some require access to niche markets that institutions avoid (hello crypto!)
- Some are noisy or slow to converge, resulting in a return profile that’s just not attractive to bigger players
The Uncomfortable Truth
Both approaches have something genuinely unpleasant about them.
Risk premia come with volatility and drawdowns. Flow opportunities are often small, intermittent, operationally complex, or all of the above.
This unpleasantness is what keeps competition at bay and allows edges to persist. If it were easy and comfortable, everyone would do it.
The unpleasantness is the opportunity. Our job is to embrace and manage it.
Final Word
Trading is not a battle of wits or a competition to outsmart the market. It’s providing services the market values: bearing risks others won’t, providing liquidity when forced/hurried participants need it, or smoothing out expected market impact.
The market will always need someone to provide these services.
And it’s in these services that the indie trader can find an edge – a trade that they expect to pay them in the long run, despite the natural variance along the way.
And it’s here that we must start. Start with the edge. Everything else is an implementation detail. And while a good implementation will help you realise maximum value from your edge, it’s not an edge in itself.
So focus on edge, and as the Whitlams say, get all the girls.
