Our Experience with TradingView Indicators, Signals and Crypto Bots

Nick (SDPF)
mintyfund

--

The holy grail of indicators, it’s said only Satoshi himself can read this.

So, you want to take your newfound love for crypto trading that little bit further and have stumbled into the world of bots, signals and TradingView indicators. Seems quite daunting at first, doesn’t it? In this guide, we’ll try to explain things as best as possible to make you feel less arrrrghhh and more ‘ahhhhh’.

Before we created our platform. minty initially started out on TradingView, sharing strategies with our Discord community and playing with pretty much all the best bots possible to use them with — we thought we’d share our two cents on what the deal was… and why we ended up porting all our strategies over to our own platform instead!

TradingView indicators

TradingView is a brilliant first step for those wanting to get a bit more insight into the world of technical analysis and more advanced charting features — not to mention they have a pretty big community of traders sharing ideas, some go as far as creating indicators for others to apply to their charts and follow their signals.

This is great until you realise an indicator you are using is rubbish. The underlying problem with TradingView indicators is the fact that ones intended for providing signals are closed-source, meaning you can’t see their code. For a developer, it’s pretty easy to re-paint indicator labels to make them seem like it’s the best thing since sliced bread, with a 95% win rate and +30% average trade. This isn’t how things work and is totally unrealistic. Indicator developers are also able to make the signals offset, meaning they usually appear 2 or 3 candles behind the most recent — in the world of crypto, this is essentially pointless if you are planning on following these signals, it’s not so bad if you are just planning on using them for confirmations though.

An easy way of checking if an indicator you are using is re-painting or offsetting is to load it on a 1-minute timeframe chart. If the signal label appears on the current candle, it doesn’t offset. If you notice the labels only seem to appear x candles behind the current, then you’ve got an offsetting indicator. For re-painting, the best way to check for this is to zoom out and screenshot where the labels are, wait a while then refresh the chart. If the signals have all moved then that’s also a red flag. They’ll even sometimes adjust themselves in front of your eyes, which is so not bueno.

Now, not all indicators do this and there are some pretty awesome ones that don’t repaint/offset and are mainly to help you confirm with other indicators/your trading plan/technical analysis if you want to enter or exit a position or not. You will need some level of knowledge around technical analysis to be able to use these efficiently. Not to mention you will also need to keep an eye on your charts most of the time so you don’t miss an opportunity! It is possible to set alerts on TradingView to fire a webhook, ping you an email/SMS or pop up on your screen when an indicators signal appears on your chart, so this helps, kind of.

TradingView strategies

There is another part of TradingView that most indicator developers tend to give a wide birth, strategies. TradingView strategies are pretty much identical to indicators in their functionality but the one key difference is the fact that you are able to back-test their signals performance over past data. Most indicator developers won’t port their indicators to strategies because to be honest, they don’t actually perform that well or weren’t intended for automated bot trading in the first place.

There are two types of developer who will create TradingView strategies, the ones who want to flaw results using re-painting and off-setting the labels to achieve the best ‘results’ and those who have some solid ideas, yet you’d still need to tinker with a LOT of indicator settings which can lead to the strategies becoming over-fit and not viable for future performance, which is a common complaint amongst the TradingView community.

We previously mentioned their alert capability too, these sound good at first, but imagine running signals from a solid strategy on a 15-minute timeframe. Fast forward an entry and exit signal later, the TradingView strategy says you made a +3% profit. Your trading bot and portfolio say otherwise… TradingView alert lag is a thing and it was the bane of our lives.

using a crypto trading bot

Before we decided to call it a day on TradingView, we had 5 public strategies — they didn’t repaint or offset and most importantly were not overfit, yet still performed as well as we’d hoped. We took a look into starting to use the alerts to fire web-hook commands to some public trading bots to start using actual funds and actively trading the strategy signals for us, completely automated and require zero manual interjection.

Things went well for a few months, however, we (and the minty Discord community) did notice some discrepancies in terms of bot performance compared to the TradingView strategy performance. Long story short, it was lag. We accounted for slippage in our strategies, but not lag.

There were cases when TradingView alerts would sometimes fire up to 2 minutes after the signal actually appeared on the chart, combine this with trading bot processing time. One bot we were using (Wunderbit) could take up to 20 seconds to actually enter/exit a trade after receiving the signal. The worst-case scenario is entering a trade 2 minutes and 20 seconds after the signal actually fired. That’s a long time in the world of crypto for obvious reasons *cough* volatility *cough*, especially if you’re wanting to run strategies on lower time-frame charts and alt-coins.

It wasn’t much of an issue for the higher timeframes, although the stats were skewed, it was more of an annoyance and completely out of our hands. The plus side was, due to the delay, sometimes closing winning positions a bit later meant that if the coin was pumping, it would actually gain more profit. However, imagine a coin pumping and a sell signal fires near the peak, the coin dumps before the signal is received and you sell at a loss… if the bot even received the signal in the first place.

We just want to mention there were other bots that processed the signals a lot faster (0–2 seconds), but the initial signal webhook alert lag still shafted them for us. Along with the occasional missed signal, it wasn’t ideal. Hence why our platform was born.

making your own bot

If you are a programming whiz or have the patience to learn a new language along with technical analysis, you can actually code and create your own indicators and trading bot in Python (for example), this essentially means you’d be able to run your own indicators/strategies connected directly to the exchange, without the need for a 3rd party like TradingView. We won’t go into detail on how to make your own bot as, to be honest, why would we when we’ve already done it and are providing the signals already! Go register and take a look, if you haven’t already.

what have we done?

Because of our issues with the above, we decided to get our heads down and port our TradingView strategies into our own platform, connect it to some exchanges directly and let them rip. Essentially cutting out the middlemen aka TradingView and bots. This means that the platform is reading the charts in real-time, exactly the same as you would be on the exchange — with our algorithms working alongside looking for entry and exit signals. When the signals fire, the platform will open and close positions almost instantly (talking mere milliseconds). Lag is pretty much non-existent now.

Not only that, as you may have already noticed, we launched the platform in paper mode (read more about that here), meaning the strategies will be running live and the results seen will be the results you will get IRL. Most importantly, there will definitely be none of this re-painting/offsetting nonsense — we aim to be as open and transparent as possible. The only thing we won’t do is share our algorithms' source code, obviously.

We are also taking things even further by implementing AI and machine learning into our algorithms. We aren’t just saying AI for the buzzword either, it’s actually happening now we aren’t restrained by the limited functions of TradingViews PinceScript language. Join our Discord and create an account on the platform to see how they will perform over the coming months once we have added them.

Thanks,
The team @ minty

--

--

Founder of https://minty.fund — crypto addict, chartered marketer and lover of all things digital.