Algorithmic Trading Gets Smarter After Quant Upset

THE-BEAKER · Oct 5, 2007

aha i spy a yellow swan with pink spots.

thats only suppose to happen once in a billion years.

i think this sums up quant funds pretty well.

Reports that say that something hasn't happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns -- the ones we don't know we don't know."

Donald Rumsfeld

gbos · Oct 5, 2007

Code:

Sub  Lets_Make_a_H_Fund()
   Dim FundName as string
   FundName = âGlobalAlphaâ
   Dim Other_Peoples_Money as Double
   Other_Peoples_Money = 10 billions
      If things = âcalmâ then 
        Other_Peoples_Money = 12 billions
        Our_Fees&profit = 1 billion
      Else
        Other_Peoples_Money = 6 billions
        Print "25 sigma event"
        Call Sub Start_A_New_H_Fund
      End if
End Sub

AK100 · Oct 5, 2007

Many of these guys aren't as stupid as everyone thinks because they know their models are crap in times of panic.

Most of that talk is pure marketing so as to give the investors the impression that they're employing pros/people who really know what they're doing.

In the meantime the fund managers know that when the crap hits the fan it's likely that their funds will take big hits but they hope that first the markets will trade normally so they can rape the fund of fees and pay themselves some big $.

So with the Goldman boys how much have they personally paid themselves over the last 5 years and can investors get any of it back? No chance.

Heads they win, tails they don't lose (I just don't make a lot of money via performance related bonuses).

THE-BEAKER · Oct 5, 2007

Quote from gbos:

Code:

Sub Lets_Make_a_H_Fund() Dim FundName as string FundName = âGlobalAlphaâ Dim Other_Peoples_Money as Double Other_Peoples_Money = 10 billions If things = âcalmâ then Other_Peoples_Money = 12 billions Our_Fees&profit = 1 billion Else Other_Peoples_Money = 6 billions Print "25 sigma event" Call Sub Start_A_New_H_Fund End if End Sub

sad thing this is truer than most people realise

mokwit · Oct 5, 2007

I agree that these quant funds are really just marketing of leverage in a bull market. Walt & Irma did welll buying the dips last time until they bought the last one.

QuantPlus · Oct 5, 2007

"Quant" is now officially a meaningless term...
Just like "hedge fund".

The fact is...
That any market-neutral fund...
That quantitatively exploits market inefficiencies...
Would have had EASY record profits in Third Quarter.

We did.

Because greater volatility ALWAYS means greater market inefficiency (or mispricing).

The hard part is to stay market neutral in a crisis...
Which is where the losers failed...
Because they were pairing up securities with marginal correlations...
And hence took big one-time losses.

makloda · Oct 5, 2007

Quote from AK100:
So with the Goldman boys how much have they personally paid themselves over the last 5 years and can investors get any of it back? No chance.

I think the Global Alpha fund is now closed to outside money and mostly consists of "internal" investors, i.e. GS partners. So any losses there cut into the partners' net worth just as bad as into any remaining 3rd party investors.

Plus, I believe it was up 100% until 12/2005. This is not an excuse for any losses, just puts the recent losses a little in perspective.

AAAintheBeltway · Oct 5, 2007

Quote from QuantPlus:

"Quant" is now officially a meaningless term...
Just like "hedge fund".

The fact is...
That any market-neutral fund...
That quantitatively exploits market inefficiencies...
Would have had EASY record profits in Third Quarter.

We did.

Because greater volatility ALWAYS means greater market inefficiency (or mispricing).

The hard part is to stay market neutral in a crisis...
Which is where the losers failed...
Because they were pairing up securities with marginal correlations...
And hence took big one-time losses.

I read the article about the BSC fiasco that led to Specter getting canned. They mentioned they had hedged with something called ABX which is apparently a mortgage index but that it went up when subprime tanked, so they got hit on both ends. Is that the kind of event you are referring to?

Pretty fascinating. I hope you can explain this in more detail.

smilingsynic · Oct 5, 2007

Quote from kinggyppo:

David Viniar, chief financial officer of Goldman Sachs â whose flagship quantitative multi-strategy Global Alpha hedge fund is down 32.9% this year to mid-September, and in August plunged 22.5% â termed Augustâs market moves as a 25-standard deviation event, something that would normally only occur once every 100,000 years. In September, the fund was valued at $6bn (â¬4.3bn), down from $10bn last year.

Is this true, the correction was 25 standard deviations?

Once every 100,000 years? LOL. If I weren't comfortable in my lounger, I'd be ROFLMAO.

Let's get real.

The NYSE opened in 1792, a little over 200 years ago. The history of financial markets doesn't go back much beyond that.

When was the first hedge fund established? Less than sixty years ago, I think, and A.W. Jones, the first hedge fund manager, wasn't much of a quant.

100,000 years ago Borg was knocking the heads of other cavemen with his club.

The ancient Greeks had a word for such arrogance (not Borg, but that of said clueless and much less sensible quant): hubris.

A hedge fund that bets against the hubris of "quant" hedge funds? Now THAT would be something. And I have the perfect name: the Icarus Fund, after Icarus, the mythological drowned in the sea because of his own hubris.

Like him, Mr David Vinier is all wet.

kinggyppo · Oct 6, 2007

Just saying that the august correction in the SP500 was not a
25 STD DEV move. Now if you are leveraged up the wazoo, ok.
The SPX corrected less than 10% from its peak. I expect some
bullshit as no one likes to lose, but I am surprised it printed.

1987 crash:

The magnitude of the crash took virtually all investors by surprise. Under the hypothesis that the Index is subjectively lognormally distributed with a 20% volatility (near its long-term average realized annualized volatility since 1928), the October 19th crash was a -27 standard deviation event, which therefore should only occur with the probability 10 to the -160th power -- a virtual impossibility.

http://www.in-the-money.com/glossarynet/1987_sto.htm