Against the herd: trader Mark Spitznagel on contrarian investing
ft.com
ft.com
Also, on average, my "punter" picks do about the same as the contrarian funds although with more wild swings up and down, so for less hair pulling I am tending to move more and more towards managed funds or industry-based ETFs with the occasional punter picks.
I certainly think that having some investments in the contrarian space makes sense as a diversification strategy, but I tend to leave those stock picks up to someone who has made it their day job.
How have they done against the S&P?
Though it almost seems like luck. If you had asked me 2 years ago, my contrarian funds were in the red.
I think the pandemic has really created both good contrarian opportunities as well as good value investment opportunities; both strategies seem to be working well at the moment.
https://news.yahoo.com/mark-spitznagel-univesa-cio-on-risk-m...
https://www.institutionalinvestor.com/article/b1nhg4w9k5hjp0...
They made 4000% on the premium paid in, not on capital. No practitioner uses such concept because it's not useful.
Imagine you insure your car worth $20000 and pay $1000 a year in premium. Your car gets totaled and the insurance pays out $10000. You are not going to say "woah my insurance paid out 1000%", as it's a meaningless number. You are going to say: "my insurance paid out 50% of the original car value".
This is the same case, but of course Mr Spitznagel doesn't mind that silly numbers are flying around his fund.
Since they were able to invest only the premium paid in, I don't see why this isn't a useful metric. If they return 4000% on every dollar I give them but I only choose to give them 1% of my wealth, then the lower aggregate returns are on me.
Dividing by a higher notional amount would be justified if the Universea funds could cause investors to lose more than their initial investment, in the same way that futures contracts have a notional value far in excess of their face value. However, I do not think the Universea funds work in this way.
I feel there's a misunderstanding of what Universa is and this keeps confusing people.
Universa is not an alpha fund. You don't give money to Universa with the aim that they'll generate steady income, or even any income. They are basically an insurance product. If nothing happens for 10 years, they'll just steadily lose money (this is expected) in the form of the premium. In fact, in absence of market crashes, the expected value of Universa is negative.
>Dividing by a higher notional amount would be justified if the Universea funds could cause investors to lose more than their initial investment, in the same way that futures contracts have a notional value far in excess of their face value. However, I do not think the Universea funds work in this way.
That's not at all how you think of a tail protection fund like Universa.
The premium is just a constraint, not a risk metric. Say your portfolio makes 10% a year and you are willing to invest 1% of that in protection. That 1% premium spending is your constraint, nothing more. But your risk metric is the capital that you wish to insure, not the premium. It then follows that you should calculate the Universa return as a % of the capital you want to protect, not the premium you spend to protect it.
Exmaple: you have assets worth $100 and you want to spend $1 on protection (Universa). You determine that in a 2020 style market stress event you'd be expected to suffer a 20% drawdown. Beyond protection efficiency analyses, you really don't care what your return on premium is, you care that for the 1% premium spent, your protection is expected to earn enough as a % of your assets that your total drawdowns become manageable. The premium just puts a ceiling on the overall insurance level you can afford for your portfolio, but again for this you need to think in terms of % return of your portfolio value.
Just like other managed funds don’t make promises but market themselves along a certain strategy. They market themselves along a different strategy.
Universa is a fund. They returned 100%+ on every dollar invested. End of story. You can think of them ‘like insurance’ or ‘like a casino’ but it’s ridiculous to pretend they are.
[0] https://en.wikipedia.org/wiki/A_Random_Walk_Down_Wall_Street
For me, I decided to follow Steve Clark's advice of "Do More of What Works and Less of What Doesn't". I invested in a bunch of different things including ETFs, managed funds, futures, individual stock picks, IPOs, etc. and I found managed funds that seemed to work better than ETFs (though 5 years is not really enough time to know for sure), so I pumped more into the managed funds that were working.
Investing only in IPOs seems to work well as well but the hardest part for me was knowing when to sell (which in my case I almost never sell and I often miss the best opportunity to cash out my IPO investments).
I am sure there are some bad managed funds, probably more bad ones than good ones. I'm certainly not saying managed funds are a better choice overall than just going "all-in" on Vanguard, but picking out a few managed funds to try out after doing a couple of hours research is also not a strategy that should be completely written off either, at least in my experience.
When allocating capital to funds you are being adversely selected against: those who can generate returns probably aren't interested in your money (there's a lot more money in the world than profitable trading/investment opportunities).
As such, I normally recommend to people that unless they are professionals and know that they are doing or HNW (say >50mil), they stick to ETFs for the bulk of their portfolio. To be clear this isn't the optimal strategy, but it's easy and it mostly works.
Is there any proof of that? Some traders of course will get very rich, just like some lottery-players.
https://www.gerd-kommer-invest.de/der-beste-fonds-aller-zeit...
If we're talking about short-term traders, yes. There's firms and traders that make money almost every day of the year.
If we're talking about investors with 4 week+ holding periods, then I agree the evidence is much more flimsy and the luck explanation (or risk premium explanation) can apply very often (or all the time).
In investing, the first thing you have an influence on is avoiding fees (go passive).
The second thing you have an influence on is time (remain invested).
I've been following yahoo.finance and it's full of articles about "technical analysis" meaning patterns people claim to see in stock movements and what they mean and predict about the future. Isn't that just like astrology? I wonder how it's possible that they are allowed to publish such junk?
It surely is "junk" as far as the value to the society is concerned but in finance/trading it's all about money and technical trading makes money.
It's not really like astrology because it has some logic. For example: if a stock was sold for $ 10 yesterday and now it jumped to $100, based today's trading volume and on historical data we can infer that a percentage of them will sell to make a profit. Of course it's not a sure thing so we will balance our trades(bets). We don't even have to win "this" bet/prophecy. We just need to win most of the bets or the highest bets.
Add to this the fact that if enough participants follow the same patterns they can move the market and make their bets a self fulfilling prophecy.
Because the market went up. Can you prove it was because of TA?
That's precisely what TA itself relies on. As the comment above mentions, the avowed rationale of TA is simply "other people use it too, so to that extent the patterns hold".
Renaissance Technologies seems to have a lot of black box mathematical models that they make a lot of money on. How is that different than TA? It's backtested, and analyzed in further rigorous means.
TA misses rigor, that's it. The problem: rigor is hard and many people are too immature to be rigorous, myself included as in many cases it means you need mathematical maturity.
It's weird, I view trading as one of the few fields where academics are more practical than being practical. I don't think the world has learned this lesson yet. I might be wrong of course, just an opinion.
Here are a few others.
Disclaimer: I'm a retail trader and investor. I'm beginning to take retail trading/investing as seriously as founding a startup as they compete very much for the same rewards in many cases [0].
I'm assuming you're talking about seeing the CEO on television or something. I'm not assuming an insider context as I don't want to think about insider trading.
I find your initial premise a tough one to accept. That people can detect lies based on their intuition. This is because I have learned in psychology classes that (1) lies are extremely tough to detect and (2) intuition is only reliable when you have practiced a lot in a consistent-rule-based environment (e.g. think Chess). While there are some lie detection patterns that outperform chance level, there are few and they don't outperform on a crazy scale.
Then again, I'll indulge in the premise because it's tough to make concrete premises and I think I see where you want to go with this. So let's do this.
From a quant standpoint, you could probably quantify "untrustworthy" speech somehow. You can backtest that algorithm. Obviously TA doesn't do this. Then again, I know little about TA because I think it is misguided.
From a trader standpoint: maybe you actually want to trade on the lies, especially when you know about them [1]. It really comes down to your form of ethics [2] and if you can make a higher return.
[0] The same rewards are:
Note: I'm a bit creative in this. When I was younger I wasn't, which is why I wasn't interested in trading/investing.
- Money
- Having freedom in your own way of working
- Getting to product market fit (aka understanding a market)
- Having an impact on society. I struggled how to achieve this with trading/investing for a long while, but the way to do it is to share what you know (I do that for free, I'm quite vocal in my friend group about it). Every decision in life has an "investment" component to it (what should I eat today? Should I buy a house? Should I go to uni? Should I date that person? What route should I take from A to B?). I like to take that grand view and hone it though financial markets.
- In both cases, you can create an environment where hardcore academics and truth seeking can have a huge financial and societal impact. If I ever get to the stage where I'm smart enough to make a lot of money due to truth, then at one point I'll make my strategies public if they have a strong "truth seeking" component to it (and provided I can isolate it from the luck factor involved).
[1] Example: look at the whole cryptocurrency space. Note: there are genuine believers, earnest hard workers (on protocols and such) and true visionairies. However, the opposite of those archetypes are also abundant in that space: hardcore cynical speculators, lazy scammers and copycats. So there are enough traders there that simply try to beat other people in a trade despite the fact that they believe it's all worth nothing and it's all bullshit.
[2] My younger self would find it unacceptable, so I'd never trade on it. Nowadays I find it acceptable to trade on "unethical things" in quite a few circumstances. In both cases, I still require that I think and feel deeply about it.
This is a really nice way of thinking about it. Thanks for sharing this.
One canonical example of the former is the bid-offer imbalance. That's because autocorrelation is low, variability is high, the signal can be encoded easily, the distribution is bell-shaped, the heteroskedasticity is low, and we have a tremendous amount of historical data to work with.
TA exists on the very opposite side of this gulf, where its profound difficulty to study using the traditional statistics/ML toolbox is conflated with no edge by many dogmatists in the industry. It's unbelievably difficult to come up with a good encoding of what's visually obvious to a human, and beyond that such encodings are very difficult to backtest because of the typical nasties (autocorrelated data and so forth).
In my strong opinion, the failure of TA is more a failure of the quant industry at present. I believe they will figure out how to monetize that edge, but it's beyond the scope of their tools and capabilities at present.
Astrologers look at the positions of the stars and planets. Astronomers look at the positions of the stars and planets.
Over the past centuries, astronomers have discovered such amazing things like the distances to other galaxies, the paths of quite small rocks in our solar system, the age of the universe, redshift, cosmic background radiation, and so on.
TA, basically looking at the numbers that come out of the market, is the same. You can do silly unscientific things with it, or you can apply a falsifiable method and make real progress.
If you go down the scientific route, you will find the same data takes you down a rabbit hole of low latency coding, linear algebra, infrastructure, financial theory (time value of money, optionality), economics, and so forth.
You can do this yourself on a spreadsheet:
- Get some data
- Calculate the 50 bar moving average
- apply some rule like "trade when it crosses this or that stat"
- See what the PnL is.
- Think about whether this is a fluke and whether it will keep working.
Somehow none of the ads and articles I see do this completely.
But yes, you can take the same data and use it effectively. Needs more than a spreadsheet, admittedly.
The TA crowd will tell you that TA is a self-fulfilling prophecy: People believe in patterns and whatnot, take profits at certain levels, buy at others, and that makes TA "true". However, if you think it through, it doesn't make any sense, because it is basically a price inefficiency. If levels and patterns were reliable indicators for predicting the future, there surely would also be a system that exploits these patterns, and thus, TA would be inefficient.
Imagine a hedge fund with enough money to "create" patterns in a market. You could have perfect setups to exploit everyone else in the market. Reminder: Generating alpha / active trading is a zero-sum game.
I enjoyed reading the free articles in this blog. They give you quite a bit of understanding of TA, active trading and all the fallacies around it: https://www.priceactionlab.com/Blog/article-selection/
[0] https://www.audible.in/pd/Safe-Haven-Audiobook/B09BDBLWP4
To be clear, a 10x-er is not good enough for this strategy to work. Those are the people who lead at SpaceX and make the impossible possible. What I am talking about is the person who is the 10x-er to the 10x-ers. The person who comes up with the idea and gets those 10x-ers into the same room and organizes them.
Obviously, the scale of that talent is hard to estimate, and "100x-er" is just a figurative term. Think of it this way - there should be only 10-20 such people among the CEOs of public companies (the other 100x-ers are musicians, politicians, authors, athletes, etc). They are easy to spot - definitionally, they do extremely well and will be in the fortune 500, shooting rapidly towards the fortune 50.
The cornerstone of this theory is that given enough talent, an individual can outperform the efforts of thousands or tens of thousands of regular people (hence think of the 100x-er as a figurative term). Just look at Musk going up against the entire automobile industry (while also building SpaceX on the side). Such people tend to find the right problems to work on, get the needed funding, hire the right people, and avoid all the possible calamities. Occasionally they also crack under pressure and smoke a joint on Joe Rogan's podcast, but they always prevail in the end.
Another way of looking at this is to look back in history. Rockefeller, Carnegie, Edison, Ford, Merrill, Walton, Schwab, etc. Would you want to bet against any one of them? So the key is to find that same group of people in today's society. I already mentioned the two obvious ones, but there should be another 10-20, at various stages of their careers. One that recently retired would be Bezos, some of the younger ones could be Max Levchin, Jack Dorsey, etc.
The important point here is to be extremely picky. Max and Jack are probably still too young and need a few more cycles before you can expect them to move mountains. Don't get me wrong, they are exceptionally amazing today. But if you're going to put a considerable amount of money on them (which you would have to if you buy into the idea that there are only 20 investable people of this caliber), you want them to be Steve Jobs-level at his prime in the 2000s, not Steve Jobs-level in the 1980s.
This is super contrarian and the difference between wins and losses will hinge on picking the top 20 versus the top 100 people. YMMV!
Hint: The contrarian here is not the thousands of people/institutions/funds buying on the float.
By your definition owning an iPhone is contrarian, because there are only 1billion people with iPhones.
To use your example, I would consider choosing to not use a smartphone to be a contrarian position while having an iPhone is not.
Also, not to nitpick, but Jobs and Gates are no longer running their companies, and Zuck is down to part-time. I don't consider either of those companies at the same level of growth potential as Apple back when Steve was running it.
Those things will corelate, because the 100xers aren't 100x because they can make anything happen, they are 100x because they choose the right battles and have the energy, know how, motivation and connections to deliver on it.
So I think the booming markets and the 100x entrepreneurs will almost always align.
I think Elon is a good example of that - a lot of coverage in the past years focused on missed production numbers and that sent his reputation into a tailspin. But anyone who actually bought and drove a Tesla knew that the cars were ahead of their time and that the rest of the country would eventually catch on.