Questions we get asked

These are the questions people actually send us before they hire us, with the answers we give. If yours is not here, write to us and ask it directly.

One thing to be clear about first. We build software. You bring the strategy. We do not design trading strategies, we do not sell ready-made systems, we do not recommend what to trade, and we never touch your money or your account. Your logic stays yours.

What we can build

Can you build custom algorithmic trading software from my own strategy rules and parameters?

Yes. This is most of our work. You give us the rules — entry and exit conditions, position sizing, risk limits, which instruments, which timeframes. We turn them into working software. The strategy is yours before we start and yours after we finish.

Can you code it so the whole thing is fully automated, end to end?

Yes. In practice that means we write:

  • every calculation your strategy needs, whether standard technical indicators or your own quant formulas
  • order management
  • risk management
  • execution straight through your broker's API, with no manual step

It also means handling the things that only show up once real money is moving. Historical candle data arrives with gaps, so we fill them. Feeds drop out, orders get partially filled, brokers throttle requests. Those cases go into the code, not into a list of things you have to watch for.

You also get a configurable .csv file, so you can set a ticker active or inactive yourself without anyone touching the code.

Can you run several strategies inside the same software?

Yes. Each strategy runs as its own module inside one application, with its own parameters, instruments and risk settings, while sharing a single data feed and a single broker connection. You can switch one off without disturbing the others.

Can you build custom backtesting for my strategies?

Yes. We build the backtest around your rules and your data rather than fitting your idea into someone else's tool. Where it matters, we build the backtest and the live system to share the same strategy code, so what you tested is what actually runs.

Brokers, accounts and connections

Can the software handle multiple accounts?

Yes. One system can place orders across several accounts, with position sizing and risk limits set per account rather than applied identically to all of them.

Can it connect to more than one broker at the same time?

Yes, and this comes up often when your own clients sit with different brokers. We build a connection layer that speaks to each broker's API separately, so your strategy logic lives in one place and only the connection underneath changes.

Can you take market data from a third-party provider but still execute through my broker?

Yes, and it is a sensible thing to ask for. Data and execution do not have to come from the same company. If your broker is right for execution but its data is unreliable, we take prices from a data provider you trust and keep orders flowing through the broker.

Can you build one system with multiple strategies, multiple accounts, multiple brokers, and a web dashboard?

Yes. This is the largest kind of system we build. It usually combines a strategy engine, a connection layer for each broker or platform, and a web dashboard where you can see what is running, manage users or subscriptions, and display whatever information you choose to show.

One thing worth saying plainly: we build the dashboard, but what you publish on it, and to whom, is your decision and your regulatory responsibility. We can build a screen that displays data. What that data is, and whether you are permitted to show it to a particular audience, is yours to decide.

Markets and data

Which asset classes can the software trade?

Forex, stocks, ETFs, futures, commodities and crypto. The real limit is not our software — it is what your broker and your data provider support. Tell us the instruments and the venues, and we will tell you honestly what is possible.

Can you automate crypto strategies, with backtesting?

Yes. Crypto brings its own problems: markets never close, so there is no daily settling point; exchanges differ in how they handle order types and rate limits; and historical data quality varies a lot between venues. All of that is workable, but it is worth designing for from the start rather than discovering later.

Can you help with the data I need, for backtesting and for live trading?

Yes. Sourcing data is one thing; making it usable is another. We help with both — finding a provider that covers your instruments and history, then cleaning what arrives: filling gaps, aligning timestamps across sources, handling corporate actions, and removing bad ticks that would otherwise quietly ruin a backtest.

What tends to go wrong

What edge cases and constraints should I keep in mind when designing an automated trading system?

This is the question we most wish people asked earlier. Almost none of these show up while you are researching a strategy. All of them show up in live trading.

Data

  • Historical candles arrive with gaps. If nothing fills them, indicators quietly compute the wrong values.
  • Bad ticks and spikes will trigger entries that would never have happened.
  • Corporate actions such as splits break historical continuity unless the series is adjusted.
  • Two data sources rarely agree exactly, so a backtest on one and live trading on another will diverge.

Execution

  • Orders get partially filled. The system needs to know what to do with the remainder.
  • Orders get rejected, for margin, for price bands, for instrument state. Silent rejection is the dangerous case.
  • Brokers apply rate limits. Send too fast and requests are throttled or dropped.
  • Order types are not identical across brokers. A bracket order on one may need building manually on another.

Failure and recovery

  • If the connection drops while a position is open, what should the system do? This has to be decided deliberately.
  • After a restart, does the system know what it already holds? If it does not, it can double up.
  • Reconnecting can resend orders that were already accepted, unless orders carry an identifier the broker can deduplicate.

Time

  • Time zones and daylight saving shift session boundaries.
  • Exchange holidays and special sessions differ by market.
  • Futures contracts roll over, and a naive price series jumps on roll day.

Backtest versus live

  • Slippage and latency are real and are usually absent from a backtest.
  • Look-ahead bias is easy to introduce by accident, for example by using a candle's close to decide a trade inside that same candle.
  • Costs matter: commission, spread, financing and taxes can change the character of a system entirely.

None of this means automation is fragile. It means the difference between something that works in testing and something that survives in production is mostly this list.

Choosing a platform

I have several strategies to automate. Which trading platform or broker suits them best?

We are happy to advise on this, and we are talking about technology rather than investments. The things we look at are:

  • whether the API supports the instruments and order types your rules need
  • how good the documentation is, and how stable the API has been over time
  • rate limits, and whether your trading frequency fits inside them
  • data quality, and what historical depth is available
  • what the data actually costs
  • whether it supports paper trading, so the system can be proven before it goes live

Tell us how your strategies behave — instruments, frequency, order types, how many accounts — and we will give you a straight technical opinion. We are not paid by any broker or platform to recommend them.

Working with us

What does a project cost?

There is no price list, and any number we gave you before understanding your requirement would be invented.

Asking what custom trading software costs is a little like asking what a car costs. A Toyota and a BMW are both cars, both will get you where you are going, and they are not the same purchase. Until we know which one you are describing, any number would be meaningless.

So this is how it works:

  1. Email contact@tradevectors.com with your requirement — a full write-up if you have one, or a short summary if you do not.
  2. We come back with questions specific to what you have described. This is usually the most useful step for both sides, because it is where a general idea becomes a definite specification.
  3. We then send you a cost estimate and a timeline.

If writing it all down is awkward, we are happy to do this on a call or a video meeting instead.

How long does a build take?

Between roughly ten days and several months, depending entirely on what is being built.

A straightforward system — one strategy, one broker, rules that are already clear — can be ready in about ten days. A larger system with several strategies, several brokers, a dashboard, or work that needs research before it can be written, runs into months.

You get the timeline together with the cost, after the questions step above.

Do I get the source code? Who owns it?

You can own the source code outright, with a full licence. It is yours to keep and yours to modify, by us or by anyone else you choose. At that point it is your own product rather than something you depend on us for.

That option costs an additional 30% of the project cost. Tell us at the estimate stage and it is priced into the quote, so nothing has to be renegotiated later.

What support do I get after delivery?

Every project includes one month of support after delivery. That covers edge cases and unusual situations that appear once the system is running against a live market — which, as the section above suggests, is generally when they do appear.

Anything genuinely outside the agreed scope is quoted separately, and only after we have discussed it with you first. No surprise invoices.

Can I see proof of your past work?

Yes. On a video call we will demonstrate systems we have built and show them working. Where we can, we will demo on the same broker or platform you are planning to use, so you are looking at something directly comparable rather than a general showcase.

We can also put you in touch with past clients for a reference.

Still have a question?

Send it to us. If we cannot help, we will say so, and if we know someone better suited we will tell you that too. Ask us directly.

Trade Vectors LLP provides technology consulting, software development and API integration services only. We do not provide investment advice, portfolio management or trading recommendations, and we do not manage client funds. Any trading decision, and any result of it, remains entirely yours.

๐Ÿ“… Book a Free Algo Consultation Call
Book Free Call WhatsApp