Meet The Hedge Fund Manager Who Thinks Math Is Overrated (When It Comes To Investing

D

dealmaker

Meet The Hedge Fund Manager Who Thinks Math Is Overrated (When It Comes To Investing) (Forbes)
Eric Chung, chief investment officer of Lighthaven Capital Management, joined the Contrarian Investor Podcast to discuss his view that math, financial models, and algorithms are insufficient when it comes to investing. “The widespread use of math in the investment management industry, while it can be helpful … I think there’s been some pretty significant over-reliance on these things,” says Chung. The hedge fund manager says he first started thinking about this topic after the implosion of Long Term Capital Management in 1998. “That was big news at the time,” he recalls. “An algorithmic strategy, the fund itself was run by a number of nobel laureates.” It relied very much on computer models and math to drive its investment process, but fell apart after the Russian financial crisis in 1998. More recently many quant-driven hedge funds were forced to shut down after the financial crisis in 2008.
 
dealmaker said:
It relied very much on computer models and math to drive its investment process, but fell apart
More...

These mathematical models fail because, at their core, they keep assuming that the market is random and the prices are normally distributed.

In other words, black swan events are mathematically impossible for them.

These jokers will never get it, and that's good for us trend-followers : each time they are betting that the market will soon return to "normal" because it's too high or too low, we burn their stop with our trend-following systems.
 
Last edited:
Tradex said:
These mathematical models fail because, at their core, they keep assuming that the market is random and the prices are normally distributed.

In other words, black swan events are mathematically impossible for them.

These jokers will never get it, and that's good for us trend-followers : each time they are betting that the market will soon return to "normal" because it's too high or too low, we burn their stop with our trend-following systems.
More...

Do you really think there are so dumb to not realise that and appropriately incorporate this notion in their models? Assumption about normal distribution of prices was questioned in the past century, it is now used only in simplified text book models as introduction to the topic ;)
 
Following price, learning about how that moves, especially in relation to the psychology playing out, is worth 10X more than anything a maths guy brings to the table (short term algos are the exception and of course work well).

The reason I believe is simple, the markets are like WILD animals, they are often very unpredicatble, they can and do strange things, and sometimes completely irrational. So trying to contain that via a formula is like trying to contain something that cannot be contained. Same with Fib ratios and all the other bullshit, traders are trying to contain price with their stupid numbers.

Take the LTCM debacle as a great example. They put their maths on price, they tried to contain that wild beast and they paid a heavy price. Yet, all the price action traders had no difficulty trading all that action because they bend and move with the action rather than putting constraints on it - this is what it should do etc. NO, what it should do is what it does and I should be following THAT.

Following price, learning about the price action is a long forgotton skill. Yes, many still do it, but many don't thinking superior maths skills and better tech is the way forward. It's not because the price action doesn't change over time. If in doubt, go to TradingView, put up a weekly chart of the Dow Jones, go back and look at the period of 1915-1940. Does that look any different to today's type of action? Nope. Same type of moves, same trends, same trading ranges, same sort of perversity (taking out the stops right before a big move etc).
 
Tradex said:
These mathematical models fail because, at their core, they keep assuming that the market is random and the prices are normally distributed.

In other words, black swan events are mathematically impossible for them.

These jokers will never get it, and that's good for us trend-followers : each time they are betting that the market will soon return to "normal" because it's too high or too low, we burn their stop with our trend-following systems.
More...

This guy gets it.
 
You knuckleheads are missing the whole point..

Math is a beautiful thing,but common sense and math should not be mutually exclusive..

LTCM was levered up 25-1 or more..That was their mistake,not relying on math. Let's not be ridiculous and dive off the deep end..Add in the correlation of collateral calls when shit hits the fan,and you have the doomsday scenario..
 
kroxobor said:
Do you really think there are so dumb to not realise that and appropriately incorporate this notion in their models?
More...

Of course they are dumb, just look at the gigantic crash of Long Term Capital Management for instance.

These idiots lost BILLIONS relying on mathematical formulas from NOBEL PRIZE WINNERS!

Enough said.
 
Thats a ridiculous statement..

have you enlightened with James Simons with your theories??


Tradex said:
Of course they are dumb, just look at the gigantic crash of Long Term Capital Management for instance.

These idiots lost BILLIONS relying on mathematical formulas from NOBEL PRIZE WINNERS!

Enough said.
More...
 
Tradex said:
These mathematical models fail because, at their core, they keep assuming that the market is random and the prices are normally distributed.
More...
That was back in 1973.

Read this: STATISTICAL CONSEQUENCES OF FAT TAILS Real World Preasymptotics, Epistemology, and Applications Nassim Nicholas Taleb

 
You must log in or register to reply here.
Back
Top