Posts
Comments
Plus one! I tried several free tools of this nature, but managed to find loopholes and self-sabotage every single time. Shelled out 20 bucks or something for freedom, and solved the problem instantly.
Thanks for sharing your experience—the third section in particular is really interesting. Relative to the central discussion: how much time and effort did you have to put in to find those edges, and do you think it was worth it? Would you encourage others to try and do something similar? Or is it more like a hobby?
Your formulation is nifty, and intuitively makes sense to me. I am feeling too wiped right now to think about it carefully, but my off-the-cuff response is that it has something to do with the fact that 'informational edge' is a much broader category than information about the actual underlying assets.
For example, a complicated day-trading algorithm is on some level a reflection of the fact that the market is missing some information. But that information looks more like 'there is a complex relationship between assets under XYZ specific conditions' than 'the EBTDA figure in Walmart's quarterly report'. I guess this is what most anomalies represent: they are more like meta-information than information.
In which case, if you had a superintelligence with near-infinite compute, I think your intuition that it could indeed beat the market is right. I don't know much about quants, but I imagine this is pretty much what they are trying to do, with the difference being that they have bounded resources.
Whoops, thanks, will fix that now.
Hi Scott,
If we take the 6th of March—the last trading day before the March 9 fall—then the market is down 12.2%, which is already in 'correction' territory, and an extremely rapid descent by historical standards.
If we take the 16th of March—the closest trading day we have to 'mid-March'—the market is already down 29.5% per cent, which is not too far off the bottom, and well and truly into 'bear' territory.
To be clear, I think time has proven you correct about the EMH (and this is easy for you to say, now that the market has stabilized).
I thought the talk about EMH being dead was weird at the time, and left a comment saying as much. I also wrote a post on March 24, which later turned out to be the bottom, saying that timing the market was a really bad idea, and the buy-and-hold forever strategy was about the best anyone could hope for. I am as surprised about the speed of the rebound as anyone! I possess no predictive powers, but I have consistently been defending boring orthodoxy.
I admire and respect you very much as a person and a thinker—seriously man, you have no idea—so I feel extremely bad if I have made you feel bad. I didn't mean to accuse you specifically of being a revisionist historian—it was more of a general vibe that seemed to be happening a lot—and although I think the passage as stated is misleading, I don't think it's deliberately so, and I've edited the post so it comes off as less accusatory.
Huh, I got that from a recent Bloomberg article which says 15:1...not sure who's right or why the numbers are so different.
Active management in the equity market, both in the U.S. and abroad, is dominant. And not by a little: Active management in the U.S. trounces passive by a ratio of 8-to-1 in dollar investments. Expand that to include the entire world, and the ratio is closer to 15-to-1. If we include fixed income in our calculations, the ratio balloons to 60-to-1.
Second, and it relates to the first, one of the other things (different time) that was pointed out was "market" is a problematic terms. Each asset that is traded has a market and that is not the same as "the market". I think this tends to be something of a problem area for people when the issue of EMH is in question.
Yes! This is another really great point. I think Noah Smith described it by saying something like, 'there is no EMH—there are an infinite number of EMHs all happening at the same time'. Which is another reason the theory is vague and unfalsifiable.
I find it helpful to think about each market in the narrowest possible terms. For example, the market for AAPL stock is likely to be less efficient than the S&P 500 market as a whole, although presumably not by much. The market for a thinly-traded stock languishing on the secondary board of the Tanzanian stock exchange is likely to be much less efficient than that. The market for a private company with no disclosure obligations is much less efficient again. By the time you get down to 'the market for the time and labour of this one guy called Richard', there are truckloads of inefficiencies and EMH doesn't really apply.
Thanks, nice post. I like 'anti-inductive markets', not least because it doesn't come with all the confusing connotations of 'efficient'.
Sure—I was testing a dual momentum strategy over the market as a whole, with a 12-month lookback period. The 'dual' refers to both absolute momentum, and relative momentum between asset classes (bonds, US stocks, non-US stocks).
I haven't evaluated it properly yet, but the signals it generated told me to stay in stocks until the end of March, at which point I ought move into bonds, just in time to miss the recovery. Over the period I've tested it so far (16 months) it has returned -4%, while my benchmark is at +18%. I am still mildly interested to see how it pans out, but I ignored the signals and am now only tracking it on paper.
Yeah, active investors are providing a valuable service to everyone else, both by exploiting genuine asymmetries, and by collectively generating a signal in the Uncle George sense. People sometimes worry about the passive revolution for this reason, but the vast majority of funds under management are still active, and human nature being what it is, there will presumably always be plenty of people willing to have a crack.
I think I agree with this point and I'm glad you raised it—you might non-trivially beat the market every once in a while, but not all the time. To be clear, I don't think of the EMH as an on/off switch, and if the post gives that impression, that's my bad. In fact, the most successful investors I know of wait, wait, wait, and then very occasionally load up when they see a major asymmetry.
But I do still have a couple of reservations. The first is that this would still show up in the evidential standards I've suggested? If you match the market most of the time, but every five years you beat it, you're eventually gonna have a pretty amazing track record, especially as you compound early successes. Note this is more or less the same frequency offered by some anomalies—for example, the momentum strategy I experimented with only outperforms during downturns, but ends up with a super impressive CAGR.
The second question is, how confident are you ('you' in the general sense, but also 'you' Vaniver if you feel inclined to answer) that knowing the difference between the edge trades and the others actually helps? If it's a very strong feeling, why trade on all the other occasions when you know you're guessing?
I think it's interesting in an academic sense that these edge trades might really exist, but if they still ultimately result in an unexceptional record because of all the usual human foibles, I'm not sure how it practically justifies the suggestion that anyone should do more of this sort of thing?
Each of the 'edge' investments has also had another side to it, and the other side didn't have the same feeling to guide me. For example, I correctly timed BP's bottom during the Deepwater Horizon crisis, but then when it recovered my decision of when to sell it was essentially random. I think most of the people who predicted COVID a week early were then not able to outperform the market on the other side, and various things that I've seen people say about why they expect a continued edge seemed wrong to me.
This is a super important point too. Any investor who is long the stockmarket on a long horizon ultimately has to get back in, which means they have to time the market twice. Personally, I was gobsmacked by the speed of the recovery (or the height of the dead cat bounce, if that's what it is). April was the stock market's best month in 30 years, which is not really what you expect during a global pandemic.
Crypto and domain names might be in a different class, for sure—I don't know how those markets work, although I'd be interested to hear more. But options contracts that involve e.g. the fortunes of the S&P 500 are not.
(I want to encourage rationalists to look for future opportunities to trade on, too. I just think those opportunities are vanishingly unlikely to come from highly liquid securities markets, and a lot of the recent talk is going to lead to misplaced effort.)
I don't think in any of the three cases in which Wai Dai made very high returns he would have had a problem with telling other people about his investment and it would make sense to behave as you suggest a Realistic winners would (I will of course never tell anyone about it, or it will become useless).
Yeah, I'm still not exactly sure about whether anomalies can persist in public (see one of my confusions below). But if the trade is based on a one-off event, like this one, it's obviously fine to tell people about it. If it's a formula for an ongoing edge that persists in [stock-picking/domain name-picking/crypto] then not so much.
Confusion: Can an anomaly be an open secret, and not defensible, and still somehow persist?
For example, I'm especially interested in the momentum anomaly. The momentum folks make a very compelling case that the behavioural econ reasons that brought it into existence are so dang powerful that it somehow persists, even though there are many papers/articles/books about it that anyone can read.
I personally found this just compelling enough to run a momentum experiment, based on a strategy that has done extremely well in reducing volatility and drawdowns in the past. Momentum strategies naturally lag bull markets, but more than earn their keep during the downturns, so this was the long-awaited test.
...and it performed almost comically badly. Either the anomaly evaporated, or the specific version I was testing evaporated, or it did badly this one time but still outperforms over multiple cycles, or it was only ever an artifact of overenthusiastic back-testing.
If open secrets can persist in public... for how long? What are the factors that lead to them ultimately not working? Is it possible that an anomaly could exist in public indefinitely?
Confusion: Can multiple traders profit by bringing the exact same information to market?
I want to say yes? At least if they don't individually have enough heft to fully price it in? That would explain how multiple traders are able to exploit the same anomaly, whereas hedge funds have to cap their fund size because getting too big hurts them.
Confusion: Is there some (incredibly weak) sense in which every person who makes a trade is bringing unique information to market?
Uncle George's opinion of the new iPhone is reflected in sales figures and aggregate reviews. But technically, he is in possession of a unique piece of private information: 'what does this one guy called Uncle George think about AAPL stock?'
This information is vanishingly close to being worthless, and a million miles way from a tradable edge, but it's not quite worthless—it still collectively helps to generate a signal. What does this mean, if anything?
You should certainly try to test any edge you think you have, and look for missing information. If you develop a good edge it's easy to make some profit, but it's always easier to lose money if you're careless. But don't give up before you even try.
Yep, I agree with all of this. I guess I way I would phrase it is that we don't start with a flat prior: we have mountains of evidence that most investors underperform, and that finding an edge is difficult. Doesn't mean it's not possible, and absolutely you should try, so long as you're taking very careful steps not to fool yourself about performance, benchmarking appropriately, etc.
On the 'open secrets' - I'm writing a big effort-post right now, and this is one of my main points of confusion - would appreciate your input once it's up.
Obviously some people have made money trading stocks. Does the EMH simply mean that less than 50% of people who trade stocks make money? That doesn't seem to support the grandiose conclusions that are usually made on the basis of the EMH.
Yeah exactly - for example, something like 90% of active fund managers (professional investors with all the bells and whistles) fail to beat their benchmark, and those that do are highly unlikely to repeat the feat the next year. It makes no difference to me that EMH doesn't cash out in some kind of precise formula—it just seems like a super useful and interesting thing to know. Sorry if we've been talking at cross-purposes!
Yep, true - hence high-frequency traders. I remember reading how one firm installed a direct cable between Chicago and New Jersey at some incredible expense to shave a few milliseconds off transmission time.
What would be an example of energy not being conserved in a closed system? Does the law of thermodynamics even mean anything?
I'm not sure what you're trying to say, so it would probably be better to just state your point plainly.
Like anything else, the EMH is useful insofar as it generates testable predictions about the world. One of the most useful predictions, as johnswentworth puts it: 'you shouldn't expect to make money trading stocks'.
The onus is on the person making an extraordinary claim to provide the evidence, not the other way around.
If you think you've found an exploit in the market you should absolutely start from the position that you're wrong, because... you almost certainly are. This is how anyone ought to behave purely out of naked self-interest—it has nothing to do with confirmation bias.
It means exactly what it says: asset prices reflect all available information. Not sure exactly what you're asking, but that alone is a revelatory and counterintuitive idea.
But ... this is nuts, right? It makes sense that a pandemic would make a videoconferencing company more valuable. It doesn't make sense for a completely unrelated company that may not have actually existed since 2015 to become more valuable because it happens to have a similar name as a videoconferencing company.
Seconding the call to check out Matt Levine's column, which has taught me that this kind of weirdness makes perfect sense, and is entirely consistent with EMH. If we have evidence that people often make these kind of blunders, and will continue to do so in future—another common one is KO (Coca-Cola Company) vs COKE (a bottler and distributor)—then even a totally bogus company really is 'valuable', in a very stupid but fundamentally accurate sense.
In short: the fact that people don't read tickers carefully is part of the information that markets are constantly aggregating.
Levine:
To stylize the situation, you can think of ZOOM as sort of a dormant corporate shell that owns one asset, the ZOOM ticker. It owns that asset more or less by accident, and it is not ownership in the traditional sense; stock exchanges assign tickers, and ZOOM can’t just go and sell its ticker. But still every now and then the ZOOM ticker becomes a valuable asset and starts generating revenue. In good-for-ZM times, you can make a couple of million dollars a day selling ZOOM, which I suspect—again, they don’t file financials—is more than ZOOM can make selling mobile-phone components.
Good examples. I'm gonna address this somewhat in the next post - premature optimisation seems like the bigger problem, but there are also relatively stable domains in which it makes sense to switch from exploring to exploiting as early in the piece as possible.
As you point out, the strategy hinges on the stability of the ground, which in turn hinges on the ability to determine said stability. I think there are some OK heuristics for this, although they mostly just look like asking 'what domain of life does this decision fall under?'
Yeah the comparison would work better with a fruit that hasn't been selectively bred - maybe jackfruit, or blueberries? Quick googling suggests it's really hard to find a fruit in which the sweetness signal hasn't already been distorted.
Thanks for taking the time to say so! It takes a fair amount of effort to bash the writing into shape, so it's really nice to know it doesn't go unnoticed.
I am curious about this. It's my impression that assets tend to become more correlated in a downturn. I'm not sure how much this, or the presence of fat tails, affects things, but their back test on at least three different countries' data mitigates my concern somewhat.
(I don't know how it applies to this model, but...) price movements are not normally distributed, and any model that assumes they are carries a major risk of blowing up. For example: during the financial crisis Goldman Sachs chief financial officer David Viniar infamously told the Financial Times “we were seeing things that were 25-standard deviation moves, several days in a row."
What are the chances that a 25-sigma event strikes your investment portfolio?
We should expect a 4σ event to happen twice in our lifetime. A 5σ event occurs about every 5000 years, or once since the beginning of recorded history. A 6σ event might have happened roughly twice in the millions of years since homo sapiens branched off from the other apes. A 7σ event comes along every billion years or so, or four times since our planet coalesced out of a cloud of interstellar dust. We pass the Big Bang somewhere around the 8σ mark. At 20σ, the number of years we’d have to wait is ~10x higher than the number of particles in the universe, etc.
(which is to say, Goldman and friends' models were disastrously, absurdly, cosmologically wrong.)
AFAIK Benoit Mandelbrot was the first to start warning people about this, and his PhD student Eugene Fama wrote his thesis on it...back in 1965! Which gives you a sense of how crazy it is that people would still try to apply normal distributions to financial markets.
Mandelbrot's book The Misbehaviour of Markets is worth a read. I've also written a summary of his ideas here, in the context of stress-testing the assumptions of the 'early retirement' movement.
I've been meaning to come back to this post to say thanks.
Before reading your post, it had literally never occurred to me to try decaf. I guess I thought it would be gross or pointless (to the extent I thought of it at all).
In the last two months, I've reduced my caffeine consumption by ~50%, while simultaneously increasing my total coffee consumption by about the same. This is basically the dream situation. I also noticed that most of the time, I don't really care whether it's a carefully prepared brew of the good stuff - I just want to sip on a cup of familiar-tasting hot liquid.
So, thanks for the suggestion!
Holy crap, I just noticed that most of the latest Fitbit models (including mine) have a built-in pulse oximeter!
This would presumably make it even easier to quickly map outbreaks at the population level, in the way discussed in the Lancet article (leaving aside privacy issues).
At the individual level, unfortunately Fitbit won't give me disaggregated data or a percentage reading. It only shows a graph of 'high' and 'low' variations in blood oxygen overnight, which I have no idea how to interpret. I believe this is due to FDA restrictions.
here's an example (not mine):
EDIT: can't figure out how to make image non-obnoxiously enormous, here's a link instead.
Does anyone know:
a) how accurate is a Fitbit pulse oximeter likely to be, compared to a dedicated device?
b) is there a way to access the data directly, and get a percentage reading?
c) is the 'high' and 'low' variations overnight in any way useful?
I love MMM but he dun goofed on this one. From his post:
Some people are just prone to this type of thinking, and I even have a few in my own life. They have warned me to gather “at least a few months worth” of nonperishable food in my pantry...
This is some of the juiciest low-hanging fruit anyone could pluck! Assuming you're going to use the food anyway, it is a zero-cost option that potentially makes a huge difference! And doing this as far in advance as possibly - rather than at the last minute, when people receive the officially-sanctioned 'licence' to panic - also helps to flatten the demand curve, and keep shelves fully stocked.
Cool, thanks. Still seems v. inconclusive... if anyone has more info, please chime in on that thread!
I am suspicious of the general advice against non-frontline health workers using masks (i.e. that anything short of a properly-fitted N95 mask is not only useless, but possibly does more harm than good). This air filter company claims that not only are masks a lot better than nothing; even home-made masks bodged out of cotton t-shirts or pillowcases catch 50-60% of virus-sized particles. Or as Naval suggested on Twitter, "billions of Asians aren't wrong."
I don't know whether this is correct, but am leaning towards wearing a mask in public (I have a cloth cycling mask with insertable filters, so I'm not keeping them from medical professionals). Has anyone looked into this?
Slightly meta: I'd love to see more LW posts along these lines! It wasn't until reading Sarah's post that I even realised that aesthetics matter; I've been thinking about it ever since, and I'd nominate it for the review if I could.
A common criticism of rationality/LW is that it is an aesthetic-based identity movement. I think this is true, but not necessarily a bad thing. Paul Graham's advice makes sense for politics, but he overstated the case: in my experience, 'trying on' new identities is a much better strategy for nudging the elephant in a desirable direction than attempting to convince it through reasoned argument.
I've noticed that some of the most useful identities to adopt are based around beauty/aesthetics (or screening out 'ugliness'). A simple example: I used to feel a tiny bit embarrassed for being so drawn to minimalism, as a lifestyle and as a design philosophy. The severe white apartments and Swedish furniture etc seem so masturbatory, but... I kind of like that sort of thing!
Now I notice that reducing visual clutter has a surprisingly large effect on my mood and productivity[1], and also reflects values that are important to me (frugality, conscious consumerism). Aesthetics are never entirely divorced from underlying value systems, so it makes sense that values shape your sense of style. The weird part is that it goes both ways: you can also create or adopt aesthetics that nudge your underlying value system!
I don't know if this strays into Dark Arts territory or whatever, but my wild hare-brained speculation is that playing with embodiment, identity, aesthetics, and other bottom-up cues that speak directly to the elephant might generate some interesting new breakthroughs in rationality (or post-rationality, or whatever you want to call it).
[1] Related: the entire field of environmental psychology, the extended mind thesis, JBP's 'clean your room' schtick.
Thanks for taking the time to delve into this!
You note that expected utility with a risk-averse utility function is sufficient to make appropriate choices [in those particular scenarios].
This is a slight tangent, but I'm curious to what extent you think people actually follow something that approximates this utility function in real life? It seems like some gamblers instinctively use a strategy of this nature (e.g. playing with house money) or explicitly run the numbers (e.g. the Kelly criterion). And I doubt that anyone is dumb enough to keep betting their entire bankroll on a positive EV bet until they inevitably go bust.
But in other cases (like retirement planning, as you mentioned) a lot of people really do seem to make the mistake of relying on ensemble-average probabilities. Some of them will get burned, with much more serious consequences than merely making a silly bet at the casino.
I guess what I'm asking is: even if Peters et al are wrong about expected utility, do you think they're right about the dangers of failing to understand ergodicity?
neat! thanks.
Question/feature request: does cross-posting automatically add a canonical URL element pointing to the original content? If not, would it be possible to do so? (Google doesn't necessarily penalise duplicate content, but it does effect search rankings etc.)
Not the OP, but as someone who uses both: in my mind, they're categorically different. Anki is for memorisation of discrete chunks of knowledge, for rote responses (i.e. deliberately Cached Thoughts), and for periodic reminders of things.
Zettelkasten helps with information retention too, but that's mostly a happy side-effect of the desired goal, which (for me) is synthesis. Every time I input a new chunk of knowledge, I have to decide where I should 'hang' it in my existing graph, what it rhymes with, whether it creates dissonance, and how it might be useful to current or future projects.
Once it's hanging in the lattice somewhere, I can reference and remix it as often as I want, and effectively have a bunch of building blocks ready and waiting to stack together for writing projects or problem-solving. It's fine if I can't remember most of this stuff in detail; it's much more of an 'exo-brain' than Anki, IMO.
Excellent write-up!
Anecdatum: I got into Zettelkasten before I knew what it was called after reading a post by Ryan Holiday circa 2013 (he recommends physical cards and slip boxes, too). It's profoundly improved my writing, my ability to retain information, and synthesis of new ideas, even though I was doing it 'wrong' or sub-optimally most of that time.
In terms of systems: I always thought using paper index cards was bonkers, given we have these newfangled things called 'computers', but your post makes a much more compelling case than anything else I've read (including the Smart Notes book, which is very good). So I'm pretty curious to give it a try.
My only major reservation is around portability and security. At this point, my (digital) slip-box is literally the single most valuable thing I own. I know Ryan Holiday uses fireproof safes etc, but it seems like it would get pretty cumbersome, especially once you have tens of thousands of notes.
I've been helping Conor and Josh out with Roam because I'm excited about the power-user features, but I'm pretty confident that any practice of this nature would be beneficial to students, researchers, and writers. Prior to Roam, I was using a mixture of Google Docs, Evernote, etc, which wasn't optimal, but still worked OK.
An important point you touched on which is worth stressing: the benefits of Zettelkasten accrue in a non-linear fashion over time, as the graph becomes more connected. So even if you 'get it' as soon as you start playing around with the cards, you could reasonably expect to reap much greater gains over a timespan of months or years (at least, that's my experience!).
For simple nouns and verbs, you could use pictures as the prompt? I find this really helpful for building memorable associations, and helps me 'taboo' English on the flashcards.
Another suggestion is to add some kind of personal connection or mnemonic device. I haven't used this myself, but it's recommended in a book called Fluent Forever, which is all about learning languages through spaced repetition.
I think I get you now, thanks. Not sure if this is exactly right, but one is proactive (preparing for known stressors) and one is reactive (response to unexpected stressors).
Strong upvoted. This is a great overview, thanks for putting it together! I'm going to be coming back to this again for sure.
Note that Effectuation and Antifragility explicitly trade off against each other. Antifragility trades away certainty for flexibility while Effectuation does the opposite.
Can you say more about this? You mention that effectuation involves "shift[ing] the rules such that the risks were no longer downsides", but that looks a lot like hormesis/antifragility to me. The lemonade principle in particular feels like straight-up antifragility (unexpected events/stressors are actually opportunities for growth).
Thanks for the feedback - much appreciated! I agree that the end isn't well supported (at least, in the post). I write for a general audience who want clear, actionable takeaways. If I cross-post something in the future, I'll think about editing it more heavily to fit the LW norms (i.e. explain rather than persuade).
As far as I can tell:
1. Be born to the right parents, in the right circumstances (not helpful, but important to acknowledge).
2. Apply yourself strategically in areas that compound (e.g. knowledge and skills, saving and investing, resistance training, networking).
3. Apply your effort wherever the yield is highest. All of these domains follow an S-shaped curve, with early exponential growth running into an upper ceiling of diminishing returns. At any given point in time, it might make sense to focus primarily on accumulating money, at another, skills and knowledge, at another, health and fitness, etc.
4. Choose goals that are complementary, so that each 'bucket' also helps to fill the others, and there's no single point of failure (or at the very least, avoid goals which conflict with one another).
5. Keep doing 2-4 forever. Even if you never hit that knee-shaped curve, a consistent and cumulative effort over time is pretty powerful in and of itself.
Absolutely. Another way of thinking about it is a punctuated equilibrium: in some domains it feels like nothing is happening for the longest time, then you suddenly experience 'overnight' success. I have noticed that I find projects with delayed or noisy feedback loops super stressful, even if I know there's a solid expected payoff waiting in the wings.
I am a fan of Marie Kondo and Peterson for the exact reason you describe, and enough people have mentioned IFS now that I'll have to check it out. What's the 'spoon' thing in reference to? This seems to be one of those LW-isms that I've missed somehow.
There are no negative consequences, because nothing happens in isolation. Obviously there'd be negative consequences if the average person did this, or if Berkhan ate an entire cheesecake every day. I'm not really sure what point you're making here.
There are no guidelines on this that I'm aware of, but it seems unlikely that the RDI scales linearly with lean body mass. Some proportion of micronutrient intake goes towards the likes of bones and organs and the brain, which is unchanged by having more muscle mass.
I'm less confident of this than I am of the opposite framing: people with a low caloric intake have to be more careful about eating nutrient-dense food.
Can't find good sources, it mostly seems to be anecdotal based on the ranges that strength athletes choose to stay in. My guess is that if you went too low, you'd know about it (stage-ready bodybuilders are in a world of pain). Also, kudos for maintaining a single digit body fat percentage - impressive!
Sure.
Reading the original study, it seems like one problem is that even though leptin returned to normal, it was out of sync with resting metabolic rate, which meant appetite was no longer linked to energy requirements. There is some suggestion that a slower rate of weight loss might have more success in changing the set point, but that's also contentious.
Neat, thanks! No worries about centering text. Footnotes would be much more valuable; especially the ability to automatically insert jump links (or display on mouse hover) rather than having to scroll up and down/open the document in two tabs.
From the post:
Something like 95 per cent of people who lose weight put it all back on. Almost every attempt is doomed to fail.
Fat people who are trying to lose weight are heroes, engaged in a struggle worthy of Sisyphus. Every conceivable force is levelled against them.
Not sure what gave you the impression I'm underestimating the odds, or the difficulty of the endeavour? That was literally the whole point of the post. If it wasn't communicated clearly enough, my apologies- I'd be interested in any feedback on which bits were confusing.
One of the points I was trying to make here is the underappreciated importance of path dependence and homeostasis (so a person who has always been thin will have a much easier time than someone who had to get thin).