Cash & Carry

I'd like to humbly apologize to all the readers of this thread, I fear in my enthusiasm for learning I've been taking this thread off ES's original focus, which is his carry systems, not my own simplistic models. My skills in modeling things in Excel are far greater than my social graces, and it was very rude of me to hijack this great thread about ES's Cash & Carry.

I went back and looked at all my posts, and only 2 were really contributing helpful comments about the Cash & Carry system and asking pertitent questions that everyone could benefit from. After I started understanding ES's system, I got carried away and laser-focused on a model I was developing that was night and day from Cash & Carry.

My apologies to everyone for pretty much all of my previous posts and especially to you ES for taking up so much of your limited free time asking dumb questions that didn't pertain to what you were trying to do here.

To try to make it up all that time you spent improving my system, I'm working on a historical performance chart of the Cash & Carry system so we can see historical drawdowns and overall performance of the basket as a single instrument. It'll probably take a few days, but I hope it makes up for my utter lack of manners.

Just a quick update,. All the variations of baskets (6) I've opened between Oct 4 - 15 are showing profit on capital appreciation alone.
Basket E is earning +$2.20 interest/day on margin & showing a UPL +$330 on a $2000 account. This basket has been open since Oct 4th.
As an aside, I had a demo account I set up on Oct 6th where I bought/sold "1" unit of each currency. I was going to use that as a baseline to track changes over time. Anyway, I forgot about it, and have just checked it.
Now, for a grand total investment of $1.26 I've a UPL of +$0.08. That's around 7% growth in less than a month...not bad

Finally got enough work done to post an update.... been working on the "addto" part of the equation to answer the basic question: What is the optimal way to "trade" your basket? Keep in mind: weighting is staying the same here, it's only the exposure we're tinkering with.

The current entry systems I'm exploring are:

1. Buy and Hold

2. Buy, then daily reinvest to keep your % balance invested the same (this ignores NAV changes due to basket price fluctuation and will invest a constant % of interest a day... this was my idea before that assumed basket price would remain stable) Heh, you can see how that assumption worked out.

3. Buy, then every day adjust exposure so that your % of NAV invested remains the same. Chaffcombe at the Oanda forums champions this idea, it has the effect of buying as the basket goes up in price, and selling more of the basket as it goes down. Ie, Buy High, Sell Low. The advantage of this method is that your chance of a margin Call is smaller relative to the other methods, since as the basket declines in price you're reducing your exposure. Disadvantages: heh, look at the chart.

4. Sell as basket goes up, Buy as basket goes down. I don't have a clear way of saying this one yet since I haven't programmed it out yet, but it's basically the opposite of #3, which I think might give interesting results. This is close to how ES changes Cash & Carry with his fence method, although what I'm thinking would be a daily readjustment probably, which means spread I pay will have to be factored in.

5. Using #1,2, and 3 with profit targets. Haven't done this either.

Please suggest ideas! Let's see if we can find an optimal model of reinvestment from comparing these different ideas.

Edit: removed my graph as I realized I had an error in it with regards to entry #3... Also, 1-3 did not take the spread you pay into account. My new graph has these changes but isn't quite ready. For those who saw the old one, adding spread doesnt affect the B&H (only paying it once in beginning), and slightly decreases the final profit of #2. #3 is not properly factoring in it's balance increasing from realized profits and losses, I'm not quite sure where its graph will end up.

This is great stuff everyone. Askel: did you post somewhere in this thread results of a backtest you did on several baskets regarding UPL movement. I thought I saw it but don't see it anymore. I think that is extremely valuable information.

Quoting sjburdi

Disliked

This is great stuff everyone. Aksel: did you post somewhere in this thread results of a backtest you did on several baskets regarding UPL movement. I thought I saw it but don't see it anymore. I think that is extremely valuable information.

Ignored

Yes I did, I've since removed that, as it had some inaccuracies that I wanted to fix. Even my last graph is slightly off kilter for any system that involves selling. My buying equations are all correct though, and I've made some fascinating observations. I'm almost done fixing the other two methods, and the results will be quite jaw-dropping, let me assure you. My spreadsheet has grown to over 8 MB now, lol.



Isotonic:

5. Data: Is it possible to feed free EOD data into a spreadsheet like Excel and use the correlation function to work out the coefficients yourself? Why pay?

I in fact did just that, why pay for what you can get for free easily? There's also a few other people who've calculated correlation coefficients on spreadsheets that are floating around the internet, but anyone can calculate them relatively easy with Excel. The true genius lies in what to do with those numbers

3. The base pair is NZDUSD. Why not NZDJPY or AUDJPY? Are the spreads worse? Harder to hedge?

I can't answer for ES's choices here, but I can divulge a little of my own findings: NZDUSD and AUDJPY have a 0.95 correlation coefficient over the last 5 years. Ie, they're VERY highly positively correlated. And with a daily interest rate differiential between them of .22%, you could almost go AUDJPY and short NZDUSD and capture that interest differiential with very little directional movement of the basket... I in fact have such a strategy in a FXgame account at the moment, and results are incredibly nice.

Hi, I'm 8 or so days into a 33 pair basket with almost 7% UPL already and 2% realised profit. I;ve just been adding to the winners and reducing the losers.

Hi Savant, I wonder if you add to 10 basic pairs another 2- say Eur/NOK short and Euro/DKK long, will it be ok interest-wise? Also, what about neutralizing the pound?

Please excuse my ignorance, but what does PF stand for? And where are the tutorials located?

Thank you in advance

its now over 10 years....alot has changed...want to know?