Reconstructing Synthetic TradingView Markets for Backtests

Summary

The document describes a trader’s attempt to reproduce a synthetic ETH/USD chart from an ETH/USDT market and a USDT/USD conversion series. Small differences in reported open, high, low, and close prices produced substantially different strategy signals and backtest outcomes. The proposed product of the two market series performed poorly in the author’s test, despite the synthetic chart appearing more profitable.

The practical lesson is that a synthetic symbol may not match a hand-built series from apparently related instruments. Data sources, aggregation, timing, price construction, and instrument details can affect chart values and signals; the post does not establish which factor caused the discrepancy. It offers no confirmed reconstruction method or evidence that the synthetic-market strategy will transfer to live execution on the non-synthetic pair. Its results are an individual backtest observation, not proof of an exploitable advantage.

Key ideas

• Small differences between related price series can lead to materially different strategy signals over time.
• A synthetic quote may not be reproduced by simply multiplying the apparent component markets.
• Backtests should account for how data sources and price construction affect the synthetic series.
• Signals generated on one market may not transfer reliably to execution on a related but distinct instrument.
• The post reports an unresolved discrepancy and does not validate a reconstruction method.

Tags

# TradingView Synthetic Market


# TradingView Synthetic Market












I have been trying to find out how exactly is a synthetic market calculated in tradingview, and where is the external information coming from? My goal is to design a bot to implement this strategy, but I need to know how to recreate the synthetic market exactly,

I backtested a strategy in tradingview on a synthetic market (BINANCE:ETHUSD), the results were much better than the market BINANCE:ETHUSDT (which is not synthetic, notice the "T" at the end), I compared both graphs, the OHLC differences are small (like random 0.05% or so) but overtime they caused the signals to differ a lot, what I am planning is to use the synthetic one for entries/exits, but actual trading will happen non-synthetic market which will make the most profits in my case.

I have tried this chart to try and recreate the synthetic market: BINANCE:ETHUSDT * BINANCEUS:USDTUSD = ETHUSD (Synthetic one)

But the backtest resulted in huge losses instead of profits, what can I do in this scenario, remember that my goal is to create a bot based on this strategy and the synthetic market, any solutions?

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.