Hi Guys, Wondering where can I find, Or how can I make it myself, a reconstructed tape. By that I mean, I want to print a price for every x orders. For example, I set the number of orders to be 1000, First we have to break the order flow into positive and negative orders. Then for every 1000 positive, or negative, orders Then print the actual price. Maybe Jigsaw has something like that. I don't know. But I'd rather do it myself. Keep it cheap, fun and ludicrous. Any idea ? The speculation behind this is about the path of least resistance. I've done something similar in the past with X_trader and Excel. Sometimes it takes 100 positive orders, To move the market by one point, Whereas, in the meantime, Only 50 negative orders, Move it by one point. "Follow the path of least resistance" But I remember not having great success with it. A similar thing can be analyzed with the cumulative delta.
This is called trade pressure, i.e. fraction of trades that printed at the bid or at the offer. The underlying assumption is that aggressive flow is longer term liquid takers. Most reasonable barred data providers will have something like this included in the dataset or you can generate it yourself from TaQ data. It is a well known pricing signal so it’s heavily exploited by HFTs, but you might get better results using it in related securities or combining it with other alphas.
Dark pool isn’t only about liquidity ? Hidden liquidity isn’t a problem if it’s the case. Because hidden or not, It takes x volume to move the price. If half the x is hidden liquidity isn’t the question.
Yup. I use a lot of flow related signals. My infrastructure is not good enough to compete with the « pro boys », so my flow signals are a bit longer term and a bit more tricky to calculate. It still gets reported and can be accounted for, just not as quickly as in lit markets. It gets even more complicated in the markets where they run periodic auctions. High frequency traders have various heuristics to adjust the flow metrics.