Sierra Chart Market Statistics

TICK Calculation is Not Weighted Based on Market Capitalization of Containing Stocks

Since the TICK calculation is not weighted, large-capitalization stocks will affect the TICK calculation to the same extent compared to smaller capitalization stocks.

We have seen some traders run a simple summation of the TICK-NYSE using an Intraday bar interval, and compare that to the S&P 500 cash index Daily chart to find a correlation. This is a flawed and unreliable method of analysis for several reasons. One reason is that in the case of the S&P 500 index, the S&P 500 is a weighted index of 500 large capitalization stocks on both the NYSE and NASDAQ. The TICK-NYSE is all of the stocks on the NYSE and there is no weighting.

The TICK market statistic is also a short-term market direction indicator. For example, if a stock is down for the day, it is still possible for that same stock throughout the day to have many upticks but still be down on the day compared to its previous closing price. There is not always a definitive and reliable correlation that can be made between the TICK market statistic and a weighted cash index. Often there will be a correlation but this is not always the case.

Another reason why this is an unreliable comparison is explained in the Explanation of TICK Positive or Negative Bias and How to Correct.

Sometimes we hear from users that the S&P 500 is way up/down, but the Sierra Chart TICK calculation, which is accurate, does not reflect this. Once again, this is a completely flawed method of analysis. The S&P 500 index can be up or down significantly in price, but this is due to the degree of price change of the stocks contained within it, the fact that it is a weighted index based on market capitalization, and it contains stocks from two different exchanges.

It is not possible to always make a direct correlation to the TICK-NYSE which does not use any weighting based upon market capitalization, and it includes all symbols on a single exchange. If you believe this correlation is valid, then instead look at the TICK-SP, which tracks the number of stocks that have traded at a higher price compared to their prior price minus the number of stocks trading at a lower price compared to their prior price, in the S&P 500. What you will find is that the reading given at the time you see the S&P 500 significantly up or down, is not an extreme reading and will appear to be inconsistent with the movement of the S&P 500 cash index. This definitively shows that making this kind of correlation is inherently flawed.

In summary, the TICK is a short-term calculation indicating the number of stocks that have traded at a higher price compared to their prior price minus the number of stocks trading at a lower price compared to their prior price. Whereas a weighted cash market index like the S&P 500, is indicating for a group of large capitalization stocks their last trade prices compared to their final trade prices the day before, using a weighted formula based upon market capitalization.

There is not always a direct correlation between the two.

TICK for S&P 500 Stocks Matches DTN IQ Feed

When calculating the TICK market statistic for the S&P 500 stocks using the DTN IQ Feed data feed and comparing that to the DTN IQ Feed calculated TICK, we see that there is a very close match as the below image shows.

You will notice that in the first chart at the top of this image, the timestamps differ from the other two charts below because the top chart is using delayed data. The timestamps differ by 15 minutes but that is beside the point with the comparison being made.

Interestingly you will see that the Sierra Chart calculated TICK from IQ Feed data has higher extremes which indicates more accurate reporting of these extremes. This tends to indicate that IQ Feed is using a sampling method whereas the Sierra Chart market statistics calculation involves every single trade from the S&P 500 stocks.

We also see a very close match to the Sierra Chart calculated TICK market statistic for the S&P 500 stocks using the Barchart data feed compared to the IQ Feed calculated TICK for the S&P 500.

Aside from the differences with the extremes, the small occasional variations in appearance most likely are due to time stamping differences with the final TICK output.

Therefore, the Sierra Chart TICK market statistic calculation has proven to be accurate and follows a standard calculation model. The TICK market statistics calculation method is very simple and there never has been a problem with it other than some rare instances of a problem with the underlying data feed.

Therefore, differences with the TICK-NYSE must be primarily due to small differences or variation in the particular stocks being tracked on the NYSE and to some extent time stamping differences. Sierra Chart uses a reliable list of symbols which lists the stocks on the NYSE to do the calculation. The source data comes from Barchart. As of this writing, approximately 3200 stocks are used in the calculation each day and this also matches the number of stocks that DTN IQ Feed shows are traded on the NYSE every day.

Explanation of TICK Positive or Negative Bias and How to Correct

This is an image of the S&P 500 cash index Daily chart with 3 overlay studies below it.

The first overlay is a cumulative summation of the average price of each 1 Minute chart bar of the TICK market statistic for the S&P 500 stocks calculated by Sierra Chart. This is overlaid as OHLC bars on to the S&P 500 cash index Daily chart. Notice how it slopes up.

The second overlay is a cumulative summation of the average price of each chart bar of the DTN IQ Feed calculated TICK market statistic for the S&P 500 stocks. This is overlaid as OHLC bars on to the S&P 500 cash index Daily chart. Notice how it slopes down. It goes in a completely opposite direction.

The third overlay at the bottom is the Cumulative Adjusted Value study on the chart of the TICK market statistic for the S&P 500 stocks calculated by Sierra Chart (symbol TICK-SP). This is overlaid as OHLC bars. You can now see some correlation at times to the underlying S&P 500 cash index.

As explained in the TICK for S&P 500 Stocks Matches DTN IQ Feed section we can see that there is a very close match between the Sierra Chart calculated TICK market statistic for the S&P 500 stocks compared to the IQ Feed TICK for the S&P 500.

We also want to point out, that the Barchart Data Service $TICK symbol data also has a positive bias when performing a cumulative summation of it over an extended period of time.

So why therefore is there a 100% difference of slope between the Sierra Chart calculated TICK and the DTN calculated TICK for the S&P 500 stocks?

Once again, the TICK is not a weighted calculation. Small-cap stocks will have just as much of an effect on the calculation as large-cap stocks. Although this is not relevant to the discrepancy and neither the nearly continuous positive or negative slope makes sense.

All of the reasons why one TICK is sloping up and the other is sloping down, are not exactly known. However, in the case of the IQ Feed calculated TICK for the S&P 500, at times where the market is closed the TICK value repeats itself and if it is a negative number (which it is in our test case), that then causes a downward slope of the cumulative summation calculation over that period of time. This is one of the reasons which has been observed.

However, one thing that can conclusively be said, is that this method of analyzing the TICK market statistic by running a simple summation of the values is inherently flawed and produces completely unreliable results and should never be relied upon.

What is the solution to the problem that we see? Clearly there is a solution with the third overlay in the image above. In this Cumulative Line for the Adjusted NYSE TICK article, it discusses a calculation that can be used on the TICK market statistic symbol to obtain a relative reading from the TICK. The result of this is the third overlay at the bottom of the chart above and you can see that it is more closely correlated to the actual underlying S&P 500 cash index Daily chart.

Recent Analysis

Based upon a couple of recent user comments on the Sierra Chart TICK-NYSE calculation being less than accurate, we are going to once again review this for the date 2017-01-04.

We are going to explain here once again, after analysis that the TICK-NYSE market statistic is completely correct and reasonable.

There are going to be several points of analysis.

Also we want to point out that it is incorrect to compare the degree of movement in the S&P 500 cash index and try to correlate that to the TICK-NYSE. For complete details about this, refer to TICK Calculation is Not Weighted Based on Market Capitalization of Containing Stocks.

1. We can see in the chart images below, that the number of issues/stocks the Sierra Chart stock data feed is reporting at the 9:30 AM open is 465. With IQ Feed it is 2283. This is an indication of the number of stocks that have been trading premarket. For an exchange that has about 3100 stocks on it, the 465 looks more reasonable than the 2283. We do not know the reason for the difference but once again we think with the Sierra Chart data feed which is provided by Barchart, that it is indicating a more reasonable value.
   
   
   
   
   At the end of the day IQ Feed is reporting 3134 stocks trading and the SC Data Feed is reporting 3109. Essentially the same which indicates Sierra Chart is tracking the correct number of stocks on the NYSE.
2. We can see from the above chart images that the number of issues/stocks trading on the NYSE at 9:30:59 with SC Data feed is 1564 on 2017-01-04. IQ Feed is reporting 2721. This is a significant difference and once again we do not know the reason for this difference but it seems as though the 1564 would be more reasonable for an exchange that has 3100 stocks.
3. We can see from the chart images below that the number of issues/stocks trading up at 09:30:59 on the NYSE is 906. The number of issues/stocks trading down is 612. Resulting in an TICK-NYSE reading of: 294 which is exactly what was reported by the symbol at the time.
   
   
4. Finally we are going to compare the S&P 500 cash index to TICK-SP. We will estimate the TICK-NYSE reading from the TICK-SP to indicate that the Sierra Chart TICK-NYSE is giving a reliable reading. In the below screenshots we show the S&P 500 TICK from SC Data and from IQ Feed. IQ feed has two different symbols for this.
   
   
   
   
   
   There is variation among these. And also note that the IQ Feed S&P 500 TICK which is based on trades on the exchange only (symbol: LI6N.Z) is actually higher among all of them. And we think this is really where the core problem is. You can see that the reading is higher when trade data is actually excluded from secondary exchanges. Our understanding is that the underlying Sierra Chart data feed does include secondary exchanges.
   
   Based upon the above values we can see there is an average upper value for the S&P 500 TICK of 116 at the 9:30 US Eastern time minute. Calculation: 116 = (171 + 76 + 100) / 3.
   
   The range for the TICK-NYSE is approximately +1100 to -1100. A 2200 point range. The range for the TICK-SP is approximately +400 to -400. A 800 point range. The value of 116 is 65% of the S&P 500 TICK range. Applying the same percentage to the TICK-NYSE gives us an estimated reading of 330 which is very close to the 324 TICK-NYSE value that Sierra Chart was indicating at this time. This validates that the TICK-NYSE reading was correct.

There definitely is nothing wrong with the underlying Barchart NYSE and NASDAQ data feed being used for the TICK and other market statistics calculations.

We know this for various reasons. The data is not delayed from observations we have done on various symbols at the NYSE open. And no user of the same real-time US equities data feed has said the data is delayed.

When examining the number of trades and volume from Intraday tick data for major symbols such at the ETF SPY and other major stocks, and comparing them to IQ Feed, there are no significant differences (within approximately .0009%).

We also know based on the TISS-NYSE which had a value of 3109 at the close of 2017-01-04 with SC Data and comparing to the equivalent IQ Feed symbol at the same time which had a value of 3134 that there is only a .8% difference in the number of issues/stocks being tracked. We are using a list of symbols provided by Barchart to determine what are the NYSE symbols. So once again we know the number of symbols being tracked is accurate.

We have also asked Barchart about their data feed in relation to the TICK differences, and they said: "As any trades made on NYSE listed securities we will report, and we are certainly not dropped packets which would be another explanation." , "Our US Equities data is from the original exchanges, and includes all trades and bid and ask messages.". They also said this about the other TICK calculations from the other providers: "My guess would be that the difference is others calculate these on only NYSE Arca trades and not across the exchange.".

Therefore, there is nothing at all wrong with the underlying Barchart data feed being used for the TICK calculations.

Differences with TICK between different providers primarily is going to be attributable to differences in the underlying data feed. What the specific reasons are for these differences, we do not know but we also know there is nothing wrong with the underlying Sierra Chart data feed. These differences definitely do not amount to meaning there is a problem with the Sierra Chart TICK calculation. But instead there is simply a difference and if anything the Sierra Chart TICK calculation is more reasonable than others.

So we can conclude the TICK calculations Sierra Chart is providing are 100 percent accurate based upon the underlying data and we know the underlying data is good. Therefore, concluding there is something wrong with the TICK is actually faulty reasoning and instead you should question what you are seeing from other providers based upon the fact that we are working with a known good data feed.

If you are not happy with the TICK-NYSE that we offer, then do not use it but understand what we are providing is accurate and your conclusion that there is something wrong with it is not based upon sound reasoning and is not supported by fact.

And after this analysis again, we can only conclude that what you are seeing from other providers is actually questionable and likely unreasonable and not accurate. They certainly can be misleading when comparing to the S&P 500 cash index. The TICK calculation provided by Sierra Chart is a much more reasonable and stable representation of unweighted trade activity on the market.

For additional information, refer to this Support Board posting.

Opening TICK Values

Sometimes we hear the comment, that the market has opened way up or down, and the TICK calculation does not reflect that. Once again, this is a flawed method of analysis.

Trading activity from an exchange, is a single stream of data. At the 9:30 AM US Eastern time NYSE open, the very first TICK value will either be +1 or -1, for the first stock traded. And then from there, it will increase or decrease as the data is received from subsequent trading activity since the very first trade.

Now what actually happens in the case of the Sierra Chart calculated TICK for the NYSE and other exchanges and indexes is that the calculation begins at the open of the premarket which is 4 AM US Eastern time.

So at the 9:30 AM open, the TICK will reflect the number of stocks trading up minus the number of stocks trading down at that point. However, since most stocks have not yet traded before 9:30 AM, you are not going to see an extreme up or down value at the 9:30 AM open. If you see this extreme with other provided TICK calculations, then right away you need to recognize that those values are indeed very likely erroneous and demonstrate very substandard and incompetent market statistics by these other providers. This is a basic fact.

Market Statistics Calculation Change April 2019

Effective April 25, 2019 a change was made with market statistics calculations where when the best bid/ask prices, both are beyond the last trade price, then the changes with those bid/ask prices are counted as trades with no volume which then contributes towards the calculations like the TICK and the other market statistics symbols. This applies to the symbols for all of the different exchanges and indexes market statistics are calculated for.

The way this works is that if the best Bid price is above the last trade price, and it has increased compared to the prior Bid price, then it is counted as an uptick. If the best Ask price is below the last trade price and it has decreased compared to the prior Ask price, then it is counted as a downtick. This counting of upticks and downticks continues until such time that the Bid and Ask are no longer fully above or below the last trade price.

This allows the number of stocks at the open which participate in the TICK and other calculations to be higher and causes greater extremes in the TICK calculation.

We have been monitoring this change and we believe that is a good change since some users point out that the market statistics calculations that Sierra Chart offers sometimes are different than other services, in particular around the daily open of the stock exchange.

Although we have consistently shown, that the calculations we have been providing are correct. We have written about this extensively.

So in effect, the reason for the change is to make the market statistics more predictive because we are including actual order activity which will eventually result in trades. And to reduce the number of questions we receive about the market statistics which actually has been the case.

Conclusion

Based upon all of the above information, what can be concluded? The Sierra Chart provided TICK market statistics calculations are accurate and reliable.

You will see differences among all of the different vendors TICK market statistics calculations.

Do the differences make one better than the other, perhaps with someone's particular view that might be the case but as we have shown it does not mean that the Sierra Chart calculated TICK is inaccurate.

Being that the TICK is not a weighted calculation, due to the nature of the calculation and due to some inevitable variations with the underlying stock data it is calculated from, this only reveals the inconsistent nature of the TICK calculation among data services when comparing it to the underlying stock indexes.

If you are not satisfied with the Sierra Chart calculated TICK market statistics, then perhaps it is more a nature of the TICK calculation in general which you should examine as a method of your analysis and the reliance that you currently have on it.