SynthesisDeFi - Anchored TWAPA simple Anchored TWAP created by Oliver Fujimori
Key Concept
TWAP is calculated by taking the average of multiple asset prices at regular time intervals across a set period. By averaging out these prices, TWAP helps smooth out short-term fluctuations, providing a more stable price representation over time.
Advantages of TWAP
Simplicity: The TWAP calculation is straightforward and computationally light, making it practical for on-chain calculations in DeFi.
Protection Against Flash Loan Attacks: By averaging prices over time, TWAP offers some protection against temporary price manipulations commonly seen with flash loans.
Uses and Benefits of TWAP
Reducing Market Impact for Large Orders: TWAP is used as a strategy for executing large orders by breaking them into smaller parts over a period, ensuring that the average execution price is close to the TWAP value, reducing the risk of price manipulation.
Minimizing Slippage: In DeFi, TWAP provides a stable price reference by averaging prices over time, making it less susceptible to sudden price changes (slippage) that can occur in highly volatile markets.
Protection Against Manipulation: TWAP prices are less vulnerable to flash loan attacks and sudden price spikes since they rely on multiple price points over a period rather than a single spot price.
Twap
Anchored TWAP with StDev Bands [MrShadow]TWAP with:
- Anchoring: Custom, Day, Week, Month, Quarter, Year (custom anchoring can be selected by dragging a vertical line through the chart)
- Standard Devation Bands
- Auto-coloring depending on the trend
WAP Maverick - (Dual EMA Smoothed VWAP) - [mutantdog]Short Version:
This here is my take on the popular VWAP indicator with several novel features including:
Dual EMA smoothing.
Arithmetic and Harmonic Mean plots.
Custom Anchor feat. Intraday Session Sizes.
2 Pairs of Bands.
Side Input for Connection to other Indicator.
This can be used 'out of the box' as a replacement VWAP, benefitting from smoother transitions and easy-to-use custom alerts.
By design however, this is intended to be a highly customisable alternative with many adjustable parameters and a pseudo-modular input system to connect with another indicator. Well suited for the tweakers around here and those who like to get a little more creative.
I made this primarily for crypto although it should work for other markets. Default settings are best suited to 15m timeframe - the anchor of 1 week is ideal for crypto which often follows a cyclical nature from Monday through Sunday. In 15m, the default ema length of 21 means that the wap comes to match a standard vwap towards the end of Monday. If using higher chart timeframes, i recommend decreasing the ema length to closely match this principle (suggested: for 1h chart, try length = 8; for 4h chart, length = 2 or 3 should suffice).
Note: the use of harmonic mean calculations will cause problems on any data source incorporating both positive and negative values, it may also return unusable results on extremely low-value charts (eg: low-sat coins in /btc pairs).
Long version:
The development of this project was one driven more by experimentation than a specific end-goal, however i have tried to fine-tune everything into a coherent usable end-product. With that in mind then, this walkthrough will follow something of a development chronology as i dissect the various functions.
DUAL-EMA SMOOTHING
At its core this is based upon / adapted from the standard vwap indicator provided by TradingView although I have modified and changed most of it. The first mod is the dual ema smoothing. Rather than simply applying an ema to the output of the standard vwap function, instead i have incorporated the ema in a manner analogous to the way smas are used within a standard vwma. Sticking for now with the arithmetic mean, the basic vwap calculation is simply sum(source * volume) / sum(volume) across the anchored period. In this case i have simply applied an ema to each of the numerator and denominator values resulting in ema(sum(source * volume)) / ema(sum(volume)) with the ema length independent of the anchor. This results in smoother (albeit slower) transitions than the aforementioned post-vwap method. Furthermore in the case when anchor period is equal to current timeframe, the result is a basic volume-weighted ema.
The example below shows a standard vwap (1week anchor) in blue, a 21-ema applied to the vwap in purple and a dual-21-ema smoothed wap in gold. Notably both ema types come to effectively resemble the standard vwap after around 24 hours into the new anchor session but how they behave in the meantime is very different. The dual-ema transitions quite gradually while the post-vwap ema immediately sets about trying to catch up. Incidentally. a similar and slower variation of the dual-ema can be achieved with dual-rma although i have not included it in this indicator, attempted analogues using sma or wma were far less useful however.
STANDARD DEVIATION AND BANDS
With this updated calculation, a corresponding update to the standard deviation is also required. The vwap has its own anchored volume-weighted st.dev but this cannot be used in combination with the ema smoothing so instead it has been recalculated appropriately. There are two pairs of bands with separate multipliers (stepped to 0.1x) and in both cases high and low bands can be activated or deactivated individually. An example usage for this would be to create different upper and lower bands for profit and stoploss targets. Alerts can be set easily for different crossing conditions, more on this later.
Alongside the bands, i have also added the option to shift ('Deviate') the entire indicator up or down according to a multiple of the corrected st.dev value. This has many potential uses, for example if we want to bias our analysis in one direction it may be useful to move the wap in the opposite. Or if the asset is trading within a narrow range and we are waiting on a breakout, we could shift to the desired level and set alerts accordingly. The 'Deviate' parameter applies to the entire indicator including the bands which will remain centred on the main WAP.
CUSTOM (W)ANCHOR
Ever thought about using a vwap with anchor periods smaller than a day? Here you can do just that. I've removed the Earnings/Dividends/Splits options from the basic vwap and added an 'Intraday' option instead. When selected, a custom anchor length can be created as a multiple of minutes (default steps of 60 mins but can input any value from 0 - 1440). While this may not seem at first like a useful feature for anyone except hi-speed scalpers, this actually offers more interesting potential than it appears.
When set to 0 minutes the current timeframe is always used, turning this into the basic volume-weighted ema mentioned earlier. When using other low time frames the anchor can act as a pre-ema filter creating a stepped effect akin to an adaptive MA. Used in combination with the bands, the result is a kind of volume-weighted adaptive exponential bollinger band; if such a thing does not already exist then this is where you create it. Alternatively, by combining two instances you may find potential interesting crosses between an intraday wap and a standard timeframe wap. Below is an example set to intraday with 480 mins, 2x st.dev bands and ema length 21. Included for comparison in purple is a standard 21 ema.
I'm sure there are many potential uses to be found here, so be creative and please share anything you come up with in the comments.
ARITHMETIC AND HARMONIC MEAN CALCULATIONS
The standard vwap uses the arithmetic mean in its calculation. Indeed, most mean calculations tend to be arithmetic: sma being the most widely used example. When volume weighting is involved though this can lead to a slight bias in favour of upward moves over downward. While the effect of this is minor, over longer anchor periods it can become increasingly significant. The harmonic mean, on the other hand, has the opposite effect which results in a value that is always lower than the arithmetic mean. By viewing both arithmetic and harmonic waps together, the extent to which they diverge from each other can be used as a visual reference of how much price has changed during the anchored period.
Furthermore, the harmonic mean may actually be the more appropriate one to use during downtrends or bearish periods, in principle at least. Consider that a short trade is functionally the same as a long trade on the inverse of the pair (eg: selling BTC/USD is the same as buying USD/BTC). With the harmonic mean being an inverse of the arithmetic then, it makes sense to use it instead. To illustrate this below is a snapshot of LUNA/USDT on the left with its inverse 1/(LUNA/USDT) = USDT/LUNA on the right. On both charts is a wap with identical settings, note the resistance on the left and its corresponding support on the right. It should be easy from this to see that the lower harmonic wap on the left corresponds to the upper arithmetic wap on the right. Thus, it would appear that the harmonic mean should be used in a downtrend. In principle, at least...
In reality though, it is not quite so black and white. Rarely are these values exact in their predictions and the sort of range one should allow for inaccuracies will likely be greater than the difference between these two means. Furthermore, the ema smoothing has already introduced some lag and thus additional inaccuracies. Nevertheless, the symmetry warrants its inclusion.
SIDE INPUT & ALERTS
Finally we move on to the pseudo-modular component here. While TradingView allows some interoperability between indicators, it is limited to just one connection. Any attempt to use multiple source inputs will remove this functionality completely. The workaround here is to instead use custom 'string' input menus for additional sources, preserving this function in the sole 'source' input. In this case, since the wap itself is dependant only price and volume, i have repurposed the full 'source' into the second 'side' input. This allows for a separate indicator to interact with this one that can be used for triggering alerts. You could even use another instance of this one (there is a hidden wap:mid plot intended for this use which is the midpoint between both means). Note that deleting a connected indicator may result in the deletion of those connected to it.
Preset alertconditions are available for crossings of the side input above and below the main wap, alongside several customisable alerts with corresponding visual markers based upon selectable conditions. Alerts for band crossings apply only to those that are active and only crossings of the type specified within the 'crosses' subsection of the indicator settings. The included options make it easy to create buy alerts specific to certain bands with sell alerts specific to other bands. The chart below shows two instances with differing anchor periods, both are connected with buy and sell alerts enabled for visible bands.
Okay... So that just about covers it here, i think. As mentioned earlier this is the product of various experiments while i have been learning my way around PineScript. Some of those experiments have been branched off from this in order to not over-clutter it with functions. The pseudo-modular design and the 'side' input are the result of an attempt to create a connective framework across various projects. Even on its own though, this should offer plenty of tweaking potential for anyone who likes to venture away from the usual standards, all the while still retaining its core purpose as a traders tool.
Thanks for checking this out. I look forward to any feedback below.
+ Time Weighted Average PriceThis is basically NeoButane's script (which should be more popular than it is) with a few additions, those being primarily plotted lines of the closing price of the previous TWAP, however I've also added the optionality of plotting a second TWAP of say, maybe a different resolution if you are so inclined. Also, you may plot shapes across the top or bottom of your chart color based on if current price is above or below the previous closing price of the TWAP, in case you might want to clean the chart up a bit and not plot the actual closing price lines.
But what is TWAP, exactly, you might be saying to yourself. If you're familiar with VWAP then you've probably a pretty good idea of what this is and how it works. TWAP is a calculation that defines the weighted average price over a specific time period. Traders use TWAP as a trading strategy, or more specifically, an execution strategy, to place large orders without excessively impacting the market price. They break down the large orders into several sets of small orders priced near TWAP. Basically it's VWAP but without the volume element, and most traders will likely use it in a similar fashion as they might use VWAP, and that is like a moving average--dynamic support and resistance.
I like to think of it as displaying a price range over a specified time period (such as a month or a week). This is why I think the closes of the previous period or two are so important. Losing the previous closing price or regaining it can often give you an inclination as to whether at least some of the next period (the one you're currently in) may be bearish or bullish.
Above is a more zoomed out view of ADA/USD
And here's the same image with just the closing price for the time period plotted.
Enjoy!
Flunki T-WAP minus MA Oscillator
Yo,
Possible the last of these for now, and mostly for the sake of completeness..
This is..
Another simple oscillator that show the difference in price between a selectable timeframe TWAP and a Moving Average of that TWAP
This is shown as a histogram.
Use numeric TWAP values for minutes (30, 60 ,720 whatever) and D, 3D, W, M for higher values
There is also a global timeframe which will set the timeframe for a global alternate timeframe (instead of current chart resolution)
On top of that is a Moving Average of the histogram value, shown as a blue / red line with an option to highlight this MA crossing zero, and an option to colour bars to this line.
The major difference between this Oscillator, and the other script (Flunki VWAP minus MA Oscillator)
(I treid to insert a link but it's invisible so it would seem, anyway.. )
is that VWAP is usually calculated daily, so there is a sharp move upon the daily close, as VWAP starts a new day. Using TWAP this does not occur, so gives smoother transitions ; also the timeframe for TWAP is selectable for additional wap fun.
Simple idea : Code open
Enjoy !
Flunki TWAP minus MTF TWAP Oscillator
Heys,
Herewith a another simple oscillator that show the difference in price between a selectable timeframe TWAP and another alternative selectable timeframe TWAP
This is shown as a histogram.
Use numeric TWAP values for minutes (30, 60 ,720 whatever) and D, 3D, W, M for higher values
There is also a global timeframe which will set the timeframe for a global alternate timeframe (instead of current chart resolution)
On top of that is a Moving Average of the histogram value, shown as a blue / red line with an option to highlight this MA crossing zero, and an option to colour bars to this line.
Simple idea : Code open
Enjoy !
TWAP TrendHere we are experimenting with using TWAP for trend analysis. It appears to work better than VWAP on lower timeframes.
TWAP is a tool used by algorithm based traders that allows them to distribute their orders throughout the day without disturbing price or having their positions known to rival traders in the market. It's similar to VWAP and serves the same function, except it lacks the volume aspect. This can be an issue when position entry/exit may be affected by slippage.
The main benefit of TWAP is deciding how to distribute orders throughout the chosen timeframe. However, just like the VWAP traders will normally use it as a moving average or target for price action to pull into.
There is an option to change the timeframe that TWAP is calculated from.
Originally by NeoButane:
Interval Volatility Bands [DW]This is an experimental study that utilizes Volume Weighted Average Price or Time Weighted Average Price calculations, Bollinger Bands, and Fibonacci numbers to estimate volatility over a specified interval.
First, the basis is calculated by selecting:
-VWAP, which has the option to be calculated using real volume or tick volume
-TWAP, which has the option to be calculated using the standard method or exponential method
Next, standard deviation from the basis is calculated and multiplied by a specified expansion coefficient. The result is then added to and subtracted from the basis to calculate the high and low bands.
There are three band calculation methods to chosse from in this script:
-Standard, which uses the default calculations
-Average, which takes a cumulative average of standard deviation
-Hybrid, which takes the maximum of the standard and average standard deviation methods
Lastly, the high and low band ranges are multiplied by Fibonacci Percentages 23.6 - 78.6.
A custom color scheme with eight default presets to choose from is included.
