Casino gambler flops :Paulson's fund lost 46 %
- Thread starter oilfxpro
- Start date
And before all that he had made like a 100MM just quietly doing risk arbitrage.
DT, the tendency for strategy developers to curve fit systems with thousands of variants and sub-optimizations is what makes that curve fitting.
You are right about 50 markets, I'm saying 60, but there would be only one scalable strategy doing multiple entries to scale in and close all contracts all using market orders.
I'd do 60 strategies with anywhere from 2,4,10,20,50,200, or 500 separate entries. (When I do this in NQ the profit factor increases exponentially as I add more entries).
Not only is the assumption that because the market is considered random that that means all strategies and the trades they produce will also be random is false.
You cannot beat the market trading thousands of strategies. It would be nice to trade 60 markets, but unless I know a value of what I'll really be doing, I'm sure institutions like Paulson's lose their edge because they're not confident in the models they have, and merger arbitrage returns have been totally abysmall for many years.
With $30 billion, I'd have 60 markets trading at max position size using strategies designed to scale into positions at market and exit or reverse all contracts at market.
Can't wait to see that one. Basics are not basics unless there really is professional acquaintance with trading terminology. Today I looked at scaling my strategies for target levels of drawdown, return on drawdown, calmar ratios, apr, but more than that I was looking at the results of my strategy than each backtest's results.
You can trade high frequency with nearly 100% daily profits if you're an exchange, the exchange will then lease the order flow book out to third party vendors that make it available to the public and offer such access to brokers for a fee.
There is one strategy I've liked in ticks and since it wins the prisoner's dilemna and brinkmanship games, I think that's one I'll have to do a little more research on.
The notion of having a computer generate thousands of systems is one approach, it's just a stupid and naieve one because Multicharts will do those optimizations for you without compiling thousands of idiot spreadsheets with unreplicable results.
On Paulson, he's made some leveraged bets to get where he is, but these losses seem to indicate a lack of scalability in their designs.
Multi-billion dollar risk manaement and stratey trading systems must have more under the hood than a fix-protocol matching engine that waits for you to submit your order.
History tells you where price was relatively to the past, so going forward when the strategy's you're trading win high percentages in long time frames and at least 1:1 win ratios at 50%+ win percentages, then the range of strateies really does start to look like the efficient frontier where there is a quantitative definition of efficiency obviously Paulson's fund's not doing.
Anyway, it wouldn't be 60 markets and 20 strategies per. That means you do not have a consistent method because you're bound to get conflicting signals and that shouldn't happen since you should trade each market with 1 strategy.
The strategy I'd use as I said would scale, but you only need 1 strategy per market, and if you're adept at coding building scalability in the model through multiple entries is always better than being dependent on a NOAH'S ARK of Strategies.
$30 billion would be 60 markets trading 20th's (5 percents) of the maximum lot size. That is, I can appreciate how that might be more scalable, but it is not going to ever be as profitable as what I'm suggesting. 60 markets with 20 separate entries at 5% of the maximum lot size.
For example, if NQ's max is 10,000, I'd want to do 20 entries of 500 contracts than risk poorer results with too many models.
Really, I've seen this suggestion elsewhere and thought I'd finally comment on that stupidity.
(Lastly it is mere conjecture that trading 2,200 strategies across 60 markets means you'll always have a profitable weekly result, because you probably won't, and until someone tries to do one of these simulations in the MC Pro Portfolio Backtester, I'd like to suggest keeping such conjectures as possibilities than what you'd get, because I do not believe this would be profitable even if you paid the cheapest commission rates in the world).
Paulson analyses macroeconomic events. When he is right, he makes $ billions. When he is wrong, he losses $ billions.
I think it's more like investing except he uses more leverage than a standard mutual fund.
I used to do arbitrage when I was investing in stocks. Its a very profitable strategy, but I normally would not short the company that is buying up the other company since as stated, you get hurt both ways if the deal does not go through. Instead I would scale in and out of the trade on the long side.
Trading in futures for me is using an edge based strategy to hopefully make more money intraday than just go long or short for a longer period of time.
I think it's more like investing except he uses more leverage than a standard mutual fund.
I used to do arbitrage when I was investing in stocks. Its a very profitable strategy, but I normally would not short the company that is buying up the other company since as stated, you get hurt both ways if the deal does not go through. Instead I would scale in and out of the trade on the long side.
Trading in futures for me is using an edge based strategy to hopefully make more money intraday than just go long or short for a longer period of time.
Lately, i have made a conscious effort not to use the word "trade" or "trading". It is completely edited out of my blog. Whether profitable or not, i consider it to be an aspect of the "casino economy". Since my interest in is the area of Forex, i especially avoid using the term "trade". Over there the word is "deal". I embrace the terms "bet", "gambling" and/or "speculation" more than any other words in the so-called "investment" jargon.
The main thing is to be a "professional"...part of the group that does not lose long term, and have a good understanding why a member of that group. Got to know the numbers and have a bit of a scientific approach. I would say a "pro" has the mindset that s/he would not even begin to risk $ on anything that didn't offer the promise of long-term earnings...having done the research and development (R & D) and due diligence. That could mean hundreds of hours...or more.
I've been studying roulette lately...a member of a couple of forums. You'd be surprised as some of the charts i've seen people post there. One went steadily up right through a million spins without looking back. It is currently considered unplayable because (even though it does not use martingale variations or negative progressions) the draw-downs are in the thousands of units; needs too big of a bankroll (BR) to maintain. I suggested the designer borrow some ideas from the trading community to make it playable.
I plan to play...IF i can be convinced something offers long-term, playable promise. Anyway, lots of similarities between these disciplines.
What you're people saying is that you prefer to trade with huge bets on high leverage, without much diversification with hope it will turn out good.
Instead of having 100s different methods, each protected on the downside risk.
Well, each of us has its own taste. If you're OK risking 50% of capital or even 80% -- go ahead and trade like that.