So isnt there any way to fix ?
May your logic and perspective are different but result must be accurate, I dont see the reason to stay because inaccuracy
Hey @OptionTrader
It seems one of the pic is set for daily timeframe and other for a 30 min time frame. Hence the difference. You will find similar difference between intraday timeframe and daily timeframe on other charting interfaces like in.tradingview.com. This is because intraday time frame candles are built using the ticks received intraday from the exchange which could have missing ticks. Daily candles are built using the bhavcopy released by exchange at EOD which has the OHLC considering all ticks for the day.
I think Dhan is building a great product and backing it with great customer support. I wouldn’t jump ship because there is no other such ship to jump into currently. I know certain elements are missing but that is expected to come soon.
5 Likes
A general question: Why do brokers always rely on 3rd parties for charting? 3rd party charting platform creates dependence for even the smallest of update. Why does no one create their own in house charting solution from scratch? This gives full control and can add absolutely any feature under the sun.
Brokers, if you really ask - they have to just provide execution capabilities on exchanges. Everything else a broking platform is enablers, including charts, data, insights, etc etc.
Not just charts, you will be surprised how much of tech, even for large brokers who claim to do tech is either outsourced or vendor managed. Coming specifically to charts - one can, in realistic terms be good at only with something. Charts are tough, it’s not easy and, at least from perspective of Dhan - we prefer to make trading products & platforms and use the best of ones available - which is TradingView and ChartIQ.
Also, lets say hypothetically we start building charts - TradingView is a team of 500+ and started 10+ years back -its going to take that much, and that long to be as good as they are today.
4 Likes
I am disappointed with the conceding tone of the previous reply to be honest, we see Dhan at a better place today than what TradingView or anybody else 10yrs ago.
That being said, the focus has to be back on the chart data quality – not the chart rendering softwares – chart rendering is also a solved problem, thanks to tradingview/gocharting/chartiq etc.
I highly would recommend to have your streaming data leads to take a look at their ingest aggregators and triggers – there is something highly unusual – even sharekhan which is older than most of today’s discount brokerage firms has the best data quality. (Granted they might not have all the bells and whistles but they do have sophisticated execution performance).
Apologies for the tone of that, just that I prefer stating things as they usually are. Broking platforms are incredibly complex and have many components.
Execution systems are different from Charting systems. Execution systems connect to exchanges for transactions, and charting systems connect with exchange broadcast. At least that’s how we have designed systems.
Our focus is a lot on building core trading and transaction features, components that help us ensure we provide seamless services even while we scale. We care a lot about reliability in everything we do - just this morning we had few users stating chart load times were high - we immediately were able to deploy additional machines to ensure we cater them well… mentioning this as topic is of charts. Can we build more products, yes - but there is already a lot on our plate. We have mentioned it here: 2023: What next to expect from Dhan
I sincerely appreciate your candor @PravinJ . One of the reasons why a community at Dhan is a good place because of that.
Any trader as an individual is highly dependent on what brokerage provides to them, will have to take what brokerage provides. Nothing unusual about it as long as what brokerage provides is transparent and is full of integrity.
However, it is also important to understand the problem at course here. We are not talking about the new features or new asks and for that matter – it is simply about what is already being provided but is not being met at good operational standards.
just this morning we had few users stating chart load times were high - we immediately were able to deploy additional machines to ensure we cater them well… mentioning this as topic is of charts.
This is a problem, sure, but given the experience already gained in terms of running the infrastructure for over a year, this should be bare minimum about increasing the machines on demand before the problem is surfaced and brought out by the clients. Infact, I would just go about it as a business need to just cater to the growth demand (more new clients, more connections. Data slowness problem seems to be addressed by having just enough compute capacity at disposal and for ready to absorb the demand.
What is missing here – if you ask me, honestly it is – Operational Excellence –
data slowness is a compute problem but data precision problem is not. Data precision quality is what is the concern here.
Many trading decisions are just made on what is seen (it is okay to see the data late by couple of seconds but not okay to see a wrong data). Please take a look at Unreliable charting data for Index options? about the repetition of this.
Looked at the post you mentioned, and @Naman has already answered this. Data Precision isn’t the case in here, the exchange systems are built in a way that the broadcast is snapshot of trades sent by the exchanges, it is not absolute data or entire data. More details explained in the post, quoted below.
Ok thanks for explaining that. One question. Are the ticks indexed, serialized in order?
How is it possible that data vendors like truedata, charting solution like tradingview, trading platform like sharekhan are in sync, almost all the time (well, I have not seen any so far)?
Its a snap - not a serialised one. Even when you compare on any platform that are charting focussed - it will be in sync with most platforms possibly 99 out of 100 times. Other factors that decide are - server locations, exchange pop, connectivity and more.
Charts have two components - one real time that is plotted, and other saved from past data. Past data is from servers - which usually will match > 99.5% of times, and live data that is plotted will be possibly matching 98-99 out of 100 times.
@PravinJ
why is there a different price for the same symbol? I used the candle and heikin ashi chart.
screen recording
https://www.screencast.com/t/iyVKY2tcReN
1 Like
It is totally Normal
The Heikin-Ashi technique uses a modified formula in order to calculate the candles OHLC:
- xClose = (Open+High+Low+Close)/4 - The average price of the current bar.
- xOpen = [xOpen(Previous Bar) + xClose(Previous Bar)]/2 -Midpoint of the previous bar.
- xHigh = Max(High, xOpen, xClose) - Highest value in the set.
- xLow = Min(Low, xOpen, xClose) - Lowest value in the set.
As a result, the last price you see on a regular candlestick chart is different from the Heikin Ashi chart.
well, I think the last price has nothing to do with HA candles, It doesn’t get modified by any formula in HA candles.
The actual TRADED LTP doesn’t get modified but what is displayed does get Modified. All because of the formula I mentioned above.
You should enable “BID/ASK” Lines to see the trading LTP.
On HA charts, you need to enable following Chart Settings to see the latest LTP.
It seems this option is missing in Dhan TV charts.
1 Like
Nice Spot @amit ! This is a great setting! This correctly solves the purpose to see real price.
Yea, couldn’t find this setting on Dhan TV so asked @HIR to use BID/ASK line as an alternative 
Recently, replicated your concern whenever you move your cursor to a different candle, you need to refresh the chart to show the correct RSI (Relative Strength Index) value.