Both 32 and 64 bit are optimized for multi-threading.
Tradestation 9.1: Portfolio Maestro
- Thread starter Wide Tailz
- Start date
Both 32 and 64 bit are optimized for multi-threading.
This is incorrect, you can easily run multiple strategies on the same symbol in Portfolio Backtester that comes with MultiCharts, and compare their results side by side. It won't let me attach a screenshot for some reason, so here's a link http://dl.dropbox.com/u/5815637/MultiCharts_portfolio_backtest.png
I'm pretty sure Portfolio Maestro must offer similar functionality, it's a comparable product.
I'm not sure I follow. While the metaphor is graphic, what makes TradeStation a tool for newbies?
"an advanced trader who makes money you can afford to hire a programmer and develop custom code, backtesting, FIX, etc."
Custom code means custom strategies or indicators used for individual trading. Strategies are existing ideas of a trader expressed in a way that a program can understand and process - essentially doing what the trader would have done anyway, but faster. Programmers almost never come up with new trading ideas, they simply translate from human language to machine language. It's up to the trader to provide precise technical specifications on what to trade and when. So, presumably, good traders will have original and successful ideas, which can then be run faster by putting them into a script. Is this not possible in TS?
Backtesting is the process of applying strategies to historical data and evaluating how they would have performed had they been run back then. Developing a custom backtesting engine is very expensive undertaking, and software makers like TradeStation and ourselves build and perfect them for years with the help of input from traders. Is TS backtesting engine inaccurate?
FIX is the Financial Information eXchange protocol (http://www.fixprotocol.org/) and it's simply a standard method of passing messages back and forth between a trading platform and a broker. Besides FIX there are other ways to pass messages, other APIs that are used by brokers such as Interactive Brokers and Patsystems. Using FIX does not improve trading results no more than speaking Italian rather than English makes someone a better poet. Is some broker that is using FIX objectively better at brokering than TS due to the protocol?
So, now that we've defined that terms above, what is the edge that you are talking about? And what makes TradeStation a tool for newbies? Do you mean that TS as a broker builds in some sorts of tools to limit their traders, or I'm just not following the thought process here.
Two points.
Point 1 regarding âitâs a bad assumptionâ. Assumptions cannot be good or bad, as they are methods for estimating certain events with varying accuracy, and may be acceptable to some but not to others. You should definitely be aware of what they are when using them, so you know the limitations of a particular method you are using.
Letâs say you are designing backtesting, and all the data feed gives you are OHLC values. How would you approximate the movements when there is no other data? Please suggest.
A possible answer is: âI wouldnât estimate at all since itâs too inaccurate â need to get more data to make it more accurateâ. Well, two things. First, sometimes there is just no other data, or the assumption may be suitable to some people â that doesnât make it bad. For example, some people use free daily data from FreeQuotes for longer-term analysis, which only gives OHLC? Our approximation in movements is better than no approximation at all.
Second, the option to use more data to make price movements more accurate. Itâs available as well with the Bar Magnifier, you can load historical tick and minute data, and use that to make backtesting very accurate.
Point 2 regarding âcost of tick dataâ. It all depends, how much data you need, what services you know about, and reliability of the tick data that you get. IQFeed, for example, provides 4 months of tick data with a standard subscription for stocks, futures, etc. That may be more than enough for some traders, but may be not enough for others. Our Forex feed provides 10 years of tick data for major Forex pairs.
As with any service there is a cost associated with it. Minute data granularity is sufficient for most traders under most conditions when they need a lot of history â itâs a compromise in terms of accuracy and price.
You mention that the assumption is bad, so donât use it, so we suggest you use tick data for maximum precision. But you say tick data is too expensive, so we are back to having to make assumptions. So, what do you propose?
He's not in the proposin' business, he's in the flamin' business...
BTW I've known the Portfolio Maestro team for quite a while, since before the TS acquisition. They're top-notch.
BTW I've known the Portfolio Maestro team for quite a while, since before the TS acquisition. They're top-notch.
You don't even want to buy your own data. Are you making enough money to hire programmers?
To show how much I LOVE Tradestation, here is a WFO report of one of my algos which Bill wouldn't put ten bucks on.
Green line is equity, yellow is price of the underlying stock. Starting account $10k......
Green line is equity, yellow is price of the underlying stock. Starting account $10k......
I have a strong suspicion that Bill is pursuing his Phd at some university. He sounds way too academic type. I actually asked him on another thread if he is a grad student, but he didn't answer me.
Many Phd students have wrong expectations about how highly sophisticated mathematical models can allow them to win over markets. Most of this gets corrected when they join a bank.
However, most Phd students don't have a condescending attitudes about others - reading Bill it seems he has. or maybe he just has trouble believing that there are many people out there who are doing fantastic work and making fantastic amounts of money without needing extremely sophisticated mathematics or FIX protocol!