I hear the “barbell” term thrown around a lot, particularly when it comes to investing. I think it often gets used incorrectly. The term originally comes from Nassim Taleb's work (Antifragile is a great read, highly recommended even if you don't agree with everything Taleb says). This post is my attempt at putting the pieces together to better understand this model.
The barbell strategy isn't about balance and it isn't about optimizing returns. It's about preventing ruin and taking advantage of asymmetric optionality.
Another way to put it: it's not about how to manage risk, it's about how to take risks without dying along the way.
So it's not balanced between two moderate options:
And it's not a little bit of everything:
It's combining two extremes, with much more weight given to preservation:
(If you’ve ever lifted a barbell, you’ll know that you wouldn’t actually want to try to lift a bar this imbalanced. So “barbell” probably isn’t the best analogy. In fact, that may be a source of confusion about this strategy — we think of barbells as being equally balanced.)
Let's break this down.
Taking risks without losing everything
The barbell strategy is a way to think about how to make decisions under uncertainty. It’s easiest to think of this as a framework for investment (but it applies to many other things as we'll explore later). When we’re investing money, we know that there is risk, but there is also opportunity, and we typically want to minimize risk and capitalize on the opportunity.
A core pillar of barbell strategies is preventing ruin. Taleb talks a lot about “ruin” — this means complete loss of a resource (i.e. money) or enough loss that you’re forced to pull out of the game.
The point of barbells isn’t about optimizing performance — this is I think a key place where people get confused. What good is optimizing returns if your decisions put you at risk of dying?
“Well,” you say, “I can have my cake and eat it too. I can get pretty good performance with minimal downside.”
Well, no, you probably can’t (in the long-term). This "middle ground" model of thinking is fundamentally flawed when operating in a world the probability of possible outcomes follows “fat" or "heavy" tails. One of the properties of fat tailed distributions is that things "farther out" are more likely than in normal distributions. These distributions show up everywhere in complex systems and in nature. There’s a lot more to say about fat tails – this presentation is a good intro to the math behind it. There are a bunch of mathematical consequences — the one that we hear most about are “Black Swans,” i.e. unpredictable large events that seemingly come out of nowhere.
Unfortunately, the middle ground approach when it comes to risk management is highly prevalent. We expect 7% per year from the stock market on average, and don't really consider there's a possibility of a big hidden downside. It feels safe, because it's been mostly safe in the past, but we're not taking into account that we can't possibly predict if it will fail badly in the future.
So, what do you do? How do you manage risk? Maybe there’s another way to think about it. Instead of trying manage risk, let’s structure our decision-making such that: 1) we’ll survive almost any random event; 2) we can profit from variability rather than suffering from it. In other words, we structure things so that we can take risks without worrying about dying. These are the two “ends” of the barbell.
The left side: Preventing ruin
Instead of trying to predict when a big event is going to happen (this requires an “intelligence” which is impossible to achieve in many complex systems), what if we assumed something will happen eventually?
Well, if we know something big is going to happen, then let’s think about what’s the worst that could happen. The worst is… I could die ☠️ (i.e. lose all my resources or lose enough that I need to stop playing). That's pretty bad. I don’t get to keep playing, and I’ve lost all future optionality until I can figure some other external way of replenishing resources.
Let’s make sure that doesn’t happen.
We want the ballast to take risks by being ultra conservative/preservative. We put much more weight on this side of the barbell (like 90%), and choose things that are rock-solid (while being ok with lower returns in the short-term – but as we noticed before, no point in optimizing performance if you’re at risk of ruin). That way, no matter what happens, even if we lose on all of our risky stuff — we still have the left side of the barbell.
The right side: Optionality
Ok. So what kinds of risks should we be taking?
Since we’re worried about survival, we shouldn’t take any risks that we could die from (where “death” is relative to the domain we’re talking about) — so we want limited downside. At the same time, we’re in an environment where we know there’s variability and surprising things will happen. Ideally, we want to not just not-die when random things happen, we want to thrive — so we want much higher upside than downside.
This type of asymmetric option is core to what Taleb calls antifragile. Something that’s fragile breaks when random things happen outside of “normal” operating ranges. Something that’s robust stays the same. Something that’s antifragile gains.
In the investing domain, the first thing that comes to my mind when I think about asymmetric options is startup investing — limited downside (you can't lose more than you invested), but much higher upside (you can potentially make 10x, 100x, even 1000x).
So we want the right side of our barbell to be much smaller, percentage-wise, than the left side, but we want it to be filled with small little bets. We're ok if we lose all the bets, but if we get lucky we'll hit a few and get to partake in outsized benefits.
That gives us our barbell, with the weights at the extreme end of the bar and heavily imbalanced between the two:
Barbells don’t just apply to investing. It’s a useful mental model that can be applied much more broadly to how we live our lives. Some examples:
Conservative/preservative (another way to look at this is reducing harm):
- Investing in friend/family relationships
- Eat less junk food (or eat less in general)
- Exercise regularly
- Clear mind of distractions and info overload (screentime, social media, etc)
- Reduce complexity in your life (less mental load)
- Acceptance (spiritual)
- Following interests deeply without an expectation of end goal
- Meeting new people
- Being kind to a stranger
- Writing publicly
- Building software
- Working with new people
- Spending time in service of others
- Book/movie/art that changes your perspective
Of course, it's not a universal model. For example, I’m not sure where having kids fits on this scale. Kids are very high upside, but not necessarily capped downside — and yet, having kids is the best thing I’ve ever done. The time commitment and lifestyle changes for raising kids is high, health problems in kids can be difficult and life-changing, kids don’t always turn out the way you want, etc. At the same time, there's the unlimited upside — unconditional love, a more meaningful and joyful life, learning from how they see the world.
There's also a seductive category of things which are low downside and easy to do, but are actually low upside. They tend to suck up time and get in the way of spending energy on things with much higher upside. In my experience, these are things like:
- Staying on top of every errand
- Immediately responding to every communication request
- Being super regular or efficient with how you use your time
- Inbox Zero and other productivity techniques (productivity management is another way of trying to smooth out uncertainty)
- Checking social media and email frequently
There are many aspects of life where I think taking the middle way, the balanced way, is the best way. Spirituality, finding contentment, juggling work and health and family and friends — I think balance is key in these domains.
And then there are domains where it's worth taking risks. Much of the surprises in life are hidden here. How we can structure our lives and our thinking to take more risks and be more surprised?