As you may have noticed P123 is continually evolving. Some projects are big, transformational, and take years (AI, Europe, for example). Many others can be knocked off in less than a week or two. We have a backlog of hundreds of these smaller projects and prioritizing them based on the biggest ROI is mostly a guess.
Is there a genuine interest in sponsoring projects to get them done first? We’ve done this privately in the past, with success. The latest one is this Added missing USA ETFs as well as Canadian and European ETFs. It took about 3 days and the sponsor paid us $1500.
The project price would depend on several factors like how popular it is, how long it takes, and what synergies it has with our current projects. We can create a list of projects and their costs and see if there’s traction. I think we can do 1 or 2 per month.
Off the top of my head, I’d be interested in sponsoring the following features, depending on price. In priority order:
Integrate IBKR short borrow fee and available share stock factors (historical and ongoing) and allow short strategies to have borrow cost formulas that can reference these factors.
Implement full set of IBKR/FoxRiver algorithmic order types and parameters for linked accounts.
Implement linked trading for Canadian and European/UK markets.
Support customizable, randomized order type allocations per account-strategy. For example, for strategy S in account A, I may want p123 to randomly assign 50% of orders to order type T1, 25% to T2, and 25% to T3. As part of this, I would request that order type be included in the account transaction log for proper transaction cost analysis.
Add VWAP price series and allow simulations to fill at VWAP.
Add support for slippage formulas and allow country-specific commissions (support stamp tax modeling for UK stocks, etc.)
Add option to automatically transmit buy orders to downstream broker as sufficient cash becomes available for non-margin accounts.
You can see these fall into two broad themes: expanding trading functionality and making simulations more realistic.
A “Go Fund Me” - type arrangement for new features, people vote with their $$$ for the features most important to them. Early funders get some perks to make it worth their while.
I like it.
Good idea !
We have no Crypto data and no crypto models. The market is now BIG. Adding the major Crypto currencies (maybe 5) should be easy. Developing a model (Data from Glassnode)? is more difficult but could be attractive.
Equal weight books (e.g. not cumulating the stocks in the book, every stocks gets same weight no matter how often in the different strategies of the book)
Relative strength line of the cap curve versus SPY in the cap curve chart…
Daily simulations using the daily rank (not the weekly)
The daily crypto trading volume today (2/27/2024) has surpassed total US stocks.
Marco @marco : I get the sense that you don’t like crypto but it is now becoming a mature market and Portfolio123 should consider adding as least the 5 largest cryptos (as suggested by Werner) to the system.
Not sure what to do with crypto data besides creating some pages to display it. But you can do that in many places. What’s our unique offering?
Or do you mean adding a crypto universe so that you can create a trading/ranking system based purely on TA? That’s analogous as doing an ETF trading system. We just added all missing ETFs as I mentioned at the beginning of this thread, and nobody cared besides the person that financed the project. In other words not many are making money creating purely TA systems.
Sure, if crypto keeps going up bc of FOMO (no other reason bc I don’t see any killer app), any “system” will be profitable. But then just buy the crypto and hold it, which is what I’ve done for whatever reason.
It’s quite a bit of work and we have to purchase the data from somewhere. How do we monetize it? Maybe once we get our ML/AI system up and running we can use it for crypto data. Then we would have a product edge.
---- end of rant ----
This thread is about ‘Sponsoring new P123 features’ and giving users what they want. We will create a list of projects and the sponsoring fee for each.
For crypto it would take about a week once we have the data. So around a $10K to sponsor it to cover $5K of dev work and 1 year of data costs (assuming the data costs $5K/y or less).
Depending on the price, I am willing to sponsor the following projects:
Formula-based slippage. Add the ability for users to create their own formulas for slippage instead of the variable slippage formula that Portfolio123 came up with.
Add a simulation endpoint to the API.
Add industry-specific FactSet factors.
Include dividends in realized transactions and Manage performance.
Enable buy-based simulations.
Add stop button for sims, screens, and ranking system performance tests.
This one is huge (leverage in book simulations). Most institutions that come here want to run some kind of long/short portfolio and you cannot combine it properly in books. You can make it work in the screener but not in books. This is a must-do in my opinion. The long/short guys 100% need this
Allow Price for Transactions to be chosen separately for buys and sells.
(Currently both buy and sell are the same at: previous close,next close,next open,ave of next high,low,2x close)
I would love to be able to download daily price data for the ticker/date of portfolio transactions. This would help see how much slippage there is for my trades.
The reason why I am asking is because without naming strategies (including DMs) I am seeing a strong correlation between high interest in strategies and subsequent underperformance.
You can do that with google spreadsheets. You need to download your list of transactions from P123, import it into a sheet then use “GOOGLEFINANCE” function to get for example Friday close price or Monday open and compare it to your execution price.
Example of function to get Friday price
=Index(GOOGLEFINANCE(SUBSTITUTE(P2, “:USA”, “”), “close”, W2-3, 0, “daily”), 2, 2)
Where:
P2 is the ticker name column
W2 is the date column (Monday) and calculated as follow:
=S2-weekday(S2)+2
S2 is the transaction date in Portfolio123. Since the date could be different than Monday (assuming Monday is a holiday), this formula makes sure W2 is always Monday so that W2-3 in the formula will reference Friday.
