[dave_sullivan]

Except it is correct.

https://blog.thinknewfound.com/2016/05/the-asymmetry-zone/

Literally, plot daily gains vs daily losses of S&P. If you only take 70% of the losses but get 70% of the gains (so multiply each win/loss % element in the series by 0.7, and apply the new sequence to a portfolio balance), you'll end up doing better than the market.

[qwrusz]

Ah, dave_sullivan welcome back.

How about this, below is the link to the data (S&P 500 daily returns), whoever is right will donate money to an education-related charity at 70% of the dollar amount donated by whoever is wrong. So for example if you are correct, I will donate $100 and you donate $70, and vice versa.

Trader? Risk taker? For charity?

Here's the link: https://fred.stlouisfed.org/series/SP500/downloaddata

Happy to chat about it more if you want. But just intuitively and quickly everyone should be able to see why one can't say an equal 70% up/down capture ratio will just beat the market...

1. let's agree the market can do whatever the hell it wants. Up, down, whatever.

2. Imagine a market is down -5% and then up 10%. According to your story and capture ratios, when this happens your portfolio is down only -3.5% and then up 7%. Right? 70% of the down and 70% of the up?

I think you will find in just this example the market beats your portfolio by more than 1% here. This is just a simple example and I am being generous. When you look at real data you will not only find a similar pattern, but your portfolio gets absolutely crushed by the market.

Maybe an even quicker intuitive answer: If a 70% up/down capture ratio portfolio will beat the market, why isn't this a huge thing and everyone sells/changes their regular full market S&P 500 ETFs to do that?

[dave_sullivan]

Your data only goes back 10 years. Mathematically, whether 70% up/down works out to beat the market depends on the period we're talking about.

> I think you will find in just this example the market beats your portfolio by more than 1% here. This is just a simple example and I am being generous. When you look at real data you will not only find a similar pattern, but your portfolio gets absolutely crushed by the market.

Here's a python script that generates random numbers to simulate the stock market. Each time you run it, you'll get a different result: http://pastebin.com/UNtDPjxd This is basically your "simple example" run many times side by side. Sometimes a hedged strategy works better; sometimes not.

The place I got this idea in the first place was a book I read in college about the history of hedging as a strategy. It noted one of the earlier demonstrations of why it's a good strategy was when a fund showed that participating in 70% of the gains/70% of the losses of the S&P beat the S&P. But which years? This matters. Unfortunately, I can't find the book anywhere. IIRC, I think we'd be talking about a stretch covering the 40s, 50s, 60s.

This is similar to how Milken pitched the junk bond -- it was originally based on a paper that showed a balanced portfolio of low rated bonds performs better than a balanced portfolio of high rated bonds (this is explained in Den Of Thieves). This was true back then because low credit ratings were so heavily discounted by market conditions (mainly, nobody wanted to buy them).

> why isn't this a huge thing and everyone sells/changes their regular full market S&P 500 ETFs to do that?

For the same reason that no one is pushing a diversified portfolio of junk bonds anymore; what worked in the past doesn't necessarily work in the future.

[qwrusz]

Thanks. For a second I thought I was chatting with someone you who knew what they were talking out. I appreciate you confirming you were basing your answer of something you read 10 years ago in a book, a book you can no longer find. The rest of your comment shows you didn't even study the basics of finance and economics either.

Dave, Hacker News has some of the smartest commentary I found on a News site. Questions on here regularly have people answering them where that person has experience, worked in that field professionally and have studied those topics in school and have degrees in them.

I realize this is the internet, expectations are low. People troll. Even if you wanted to help, why not wait a minute before answering to see if someone more informed or more experienced, than a book read 10 years ago will have answer. It's more a courtesy to the person asking as they get a better answer, and it stops you from looking like an idiot. Win-win.

[dang]

This comment crosses into personal attack, which is not allowed on HN regardless of how wrong someone is. Please don't post like this.

[qwrusz]

My apologies. You're totally right, dang. Thanks for keeping this site civil and all the other work done for HN.

[FabHK]

Just wanted to add that Dave's notion is not quite as absurd as you make it out to be.

As you might know, daily rebalancing leveraged ETFs (say, 2x or 3x) are basically a bad idea for long-term investments, because they are (simplistically) short vol. So, they might outperform the (non-leveraged) index/ETF if vol is low and total return over the period is high. But typically, with normal or high vol, or over long periods, they underperform non-leveraged.

By analogy, a less-than fully invested ETF/trading strategy that's basically "0.7x" leveraged will beat the (non-leveraged) index/ETF in certain trading regimes (where total return over a period is negative or modestly positive, while vol relatively high).

[dave_sullivan]

Yeah, sorry, I still think you're wrong. And you keep addressing me by name, do we know each other?