This project emerged from a personal need to stay updated about financial markets without the influence of biased viewpoints or the time-consuming nature of manual research. Tendenz aims to be a starting point for further investigation.
This is accomplished by improving upon the biggest winners/losers by considering historical volatility in combination with the most recent price movements. Specifically, Tendenz aims to rank the price movements of financial assets by their statistical significance as measured by the z-score.
By applying this methodology across an entire market, we can identify assets undergoing highly unusual events. This gets even more interesting once we extend our method to more obscure assets which typically get very little 'mainstream' coverage. Japanese rice futures collapse? Credit default swaps of a Brazilian bank shot up?
Being a first-stage MVP, which is focused on UX, IU, and PMF, this is currently only done for daily returns of US equities.
For those without the time budget for manual research and lacking interest in the opinions of journalists: Gain a traders' market insight in 5 minutes on tendenz.
fly proxy 5432 -a tendenz-db
In server/
npm i
npx prisma generate
npm run dev
NODE_ENV === 'production'
npm run build && node ./dist/jobs/...js
DATABASE_URL="postgres://{username}:{password}@localhost:5432/tendenz_server?connection_limit=5"
/tendenz_server
specifies the database of the backend serverconnection_limit=5
fixes an issue with high CPU count machines. The default setting sometimes opens more connections than postgres can handle.POLYGON_API_KEY1="XXXX"
POLYGON_API_KEY2="XXXX"
(Can be identical as long as multiple jobs don't run concurrently.)
In web/
npm i
npm run dev