I am finally ready to automate a futures strategy and am exploring different ways to do so. I am not a developer and currently use Excel (without macros) for different parts of the strategy. My goal is to eventually be fully black box, knowing that iterating from gray box will most likely be my path.
The variables I'm weighing are:
- Upfront costs
- Time to market
- Degree of independence from 3rd parties
For example, I could hire a developer to get exactly what I need, though the upfront costs and necessity for an ongoing relationship would be significant. On the other hand, I could achieve "gray box status" relatively quickly doing this myself, in Excel.
My question is, what are the LIMITS of Excel? From my research it seems to be something most traders have moved on from. Is this because the tools available to non-coders today are more powerful, or something else? Additionally I'm looking into options with more features than Excel, such as TradeStation, TT's ADL, Sierra Chart's Sheets etc.
I'm trying to weigh all options before committing my time and energy to a course. All perspectives are appreciated.
Regards.
Can you help answer these questions from other members on futures io?
I have used Excel for basic data crunching quite a bit, using no macros or VB. I've used a bit of pivot tables, creating dashboards, etc. But I have never connected it to any live market data, and am far from an excel expert.
I think as to your question... it just depends on what you're trying to do? If you are just crunching basic info using OHLC prices, or something pretty simple, Excel can be fine.
But if you're trying to connect excell to live data, that might be more of an issue, and it may be easier to just use a typical trading platform.
SO, I think you might need to provide a bit more info on what you're trying to do.
So you may or may not know that it is fairly easy to have real time equity, option and futures data populate into an excel spreadsheet, by using the export to Excel from TOS and Pasting into a blank Excel spread sheet.
The way it was explained to …
apparently the data that are imported into excel are not realtime so I am not sure you can actually trade in realtime using excel.
If you like to trade through excel I suggest to look into sierrachart, that uses a spreadsheet system to implement trades, it doesn't need any programming knowldge.
I've quite a bit of experience with excel and google sheets, there was a time not too long ago where I had programs running to scrape option chain data from NSE website to do my own calculations on it.
But to keep long story short so you can get more precise answer to your question >>
While excel and google sheets are good for moderate data processing, maybe a good EOD tool or a tool if you are okay with delayed processing. But its a quite a headache if you make it interact with live data, be it exporting directly from platform or use API to route orders to platform from excel.
And no jokes, I've spent weeks figuring out what went wrong and how to solve it. But maybe I had too much of a complicated system on excel.
Now compare that to a trading platform like Trading view, I can achieve much more with much less effort.
I think technology has just moved on, I feel its just case of much better options being out there and ease of access for non-programmers and data scientists that has changed fortunes for excel for worst.
Choice is still yours though, till the day you feel you need something better than excel, you are perfectly fine even if you use excel. Trading is very private affair.
The following user says Thank You to LastDino for this post:
Back around 2008 (or so), I used Excel with eSignal. Bringing in real-time, streaming data to Excel was no problem. Automating order entry/exit was no problem. But you had to be able to code in VBA to pull it off. Biggest negative was the price. It cost a small fortune (to me, anyway). I think I was paying something like $120/month for real-time future's data and something like an additional $60/month to use their API. Not sure what they charge today (probably much less due to the additional competition), but knowing eSignal, it still probably won't be cheap.
I would never go that route again, though. As far as I'm concerned, the best bang-for-the-buck is Sierra Chart/ACSIL... especially if it's a complicated trading strategy (not that you can't build anything complicated in Excel).
Thank you all for your responses! I've been looking into this for over a week and information overload is happening.
I am indeed planning on running functions on streaming data and sending orders directly from Excel. Sending orders from Excel is possible, though the process is different depending on the provider one is using. Given I'd be running this strategy on multiple products it seems like Excel issues could arise quickly, and be a pain to find and deal with if they can be addressed at all. Currently I'm testing out Rithmic/Excel on their free trial.
Support for Excel so far seems soft, with most saying look elsewhere (+2 for Sierra Chart).
I appreciate the input.
The following user says Thank You to Yeuts for this post:
I would like to add that many systematic traders are not programmers themselves. Most platforms offer pseudo programming languages that are very easy to understand.
Concerning programming, I think that people that have never programmed in their life, think it's a big deal....it's not.
I am a telecom engineer and I programmed in Java and C in the past. Basically you take 2 or 3 courses at university and you learn. It is really easy, so you can learn it in a couple of months. Basically it's trial and error.
So don't be too afraid to learn. I learnt when there was no YouTube and not many online resources, nowadays it will be easier and faster.
Edit: I think it is excellent for any trading task, automated or not, but it does involve some definite programming, and there is a learning curve for the platform and its programming requirements.
When one door closes, another opens.
-- Cervantes, Don Quixote
The following 3 users say Thank You to bobwest for this post:
If you are a primarily left-brained individual, I'd say go for it. I would also agree with @SBtrader82, you could probably be up and coding (somewhat) in a few months, and then it's a matter of how much time you want to invest in order to get it all up and running.
I looked at the SC/Spreadsheet system for trading and didn't like it; it felt too limited to me. However, there are a few SC/Spreadsheet gurus around here who could give you better feedback on that.
And keep this in mind: If you hire a programmer... it could get expensive... and you'll probably want that person to work under an NDA... and the never-ending back-and-forth it will take to get what you want... and then you find out the programmer doesn't really know what they claimed to know... and then you have to start all over again... and, well, by the time you go through all that, you could have been well on your way to coding it yourself.
@TremblingHand has gone through a lot of trouble putting together a programming guide for beginners wanting to learn SC/ACSIL. Read through that and... if you wake up with night tremors... hire a programmer!!
Please post any questions and problems or suggestions you have in this thread about my Sierra Chart ACSIL for Beginners to keep that thread reasonably concise and clean of tangents.
Thanks.
The following 2 users say Thank You to ondafringe for this post:
IMO, people leave Excel because they want to try a different strategy, or try modifying their strategy.
With spreadsheets, it's easy to paint yourself into a corner, regarding the structure of the sheet. The "easiest" way out of it is to start from a fresh sheet. Doing that can get old quickly.
Coding strategies in a spreadsheet is restrictive, coding in a language allows for more abstractions. The physical location of 'cells' becomes irrelevant.
Using Excel VBA can help somewhat.
Some leave because Excel becomes too slow for them, or they just desire a faster implementation.
After a certain number of rows/columns, Excel may choke.
Going with a platform like NT will be faster to implement, but will cost more. The other extreme is Python. It'll be slower to implement, but basically free ... if you can code.
Also, if you ever what to explore machine learning, you're pretty much relegated to using Python (though other languages can do some ML).
Everyone can take coding courses. Everyone can program a 'hello world' "app." But not everyone will be able to code something worthwhile.
Have you defined what the technical abilities for your tool (whether Excel or not) should be? I didn't see anything about trading for instance overnight in different markets or about the number of trades, etcetera. I'm asking because Excel is not the most stable product out there if your program would need to receive and/or send out lots of data in a very short time frame. Think "choking". It can do an awful lot but it depends on how you operate it. Day trading on your pc for a couple of hours per day is a very different ball game than trading overnight that may require a server setup with optimized uptime. These things do matter.
If you're a light trader and want to invest in yourself going forward you could either learn C# or Python, or take a much easier and cheaper path by choosing a decent open source database like MySQL, put it in the cloud for a few bucks per month, and see how far you get. There's a zillion forums with technical help and IF you have to hire someone at least it's cheapish. Another option could be using TradeStation which is an excellent broker, has all data incorporated, and comes with its own free EasyLanguage code, which from what I've seen is more than a very decent language, supported by quite a community. And if you don't like it in the end a switch to MultiCharts which basically uses the same language is easy.
After having done a lot of research myself about platforms (you may want to read the 2 threads I created) I might go the open source route myself as not to be locked in for any tool of which I won't use 80% or more anyway. If you're a very structured thinker I'm sure you're able to learn how to write basic queries and code within days. The rest comes with playing and investing in yourself before investing in markets.
