Tick sizes[1]
Executive summary
When MiFID-II is implemented in Q1 2017, the tick size rules will be governed by ESMA. The same minimum-rules will apply on all EU markets. The rules will rely on metrics rather than discretionary decision by the venues and/or the members.
Two different models are discussed by ESMA (see details below) and both of them would result in a decrease in tick size for many Nordic stocks (>200 stocks). There is still room to discuss and influence some of the parameters in the proposed models in order to reduce the most negative consequences of a decrease in tick-size.
This paper provides arguments for a conservative implementation of the parameters for a new tick size regime compared to what is proposed by ESMA.
Background
ESMA will draft regulatory technical standards (RTS) specifying minimum tick sizes for equities, ETFs and other similar instruments. Since ESMA will be responsible for setting the minimum tick size, the primary exchanges (and the market participants) will in practice have no influence over tick size rules, once they are implemented (the NCAs may have possibility to do an ad hoc change). The rules for setting tick size will apply quantifiable metrics rather than the discretion of the venue and/or the members.
Those who today argue for smaller tick-sizes, have the opportunity to avoid the spread cost by using dark pools. This possibility will be restricted in the future due to the volume cap on the pre-trade transparency waivers (the negotiated trade waiver and the reference price waiver) of 4% and 8% for one market and for the total trading in the Union respectively. Furthermore, the “Trading obligation” will most likely concentrate more of the trading to the order books. This will probably increase the pressure from those who want smaller tick sizes.
In ESMA’s consultation on MiFID II, one can read that there is a target spread of around 6-7 bp.
· “Moreover, looking at market participants, it appears that when the spread on this blue chip is between 6 bps and 7 bps, their activity is well balanced between aggressive trades and passive trades”. MiFID-2 Discussion Paper ESMA/2014/548 p 308, item 81.
In order to reach the target of well -balanced tick sizes, a new definition “Spread-to-tick-ratio” is introduced (see below).
Definitions
· Spread-to-tick-ratio (STTR) – “number of ticks between the bid and the offer”
- STTR = [Ask-Bid] / [tick size] weighted over the day
- Assume a stock with tick size = 0,05 (e g VOLV B).
§ If the stock has a price that lies on minimum tick size the whole day, STTR will be 1.0.
§ If the stock have a spread of 1 tick 60% of the time, and spread of 2 ticks 40% of the day, the STTR will be 1,40.
§ The table below shows the STTR for different combination of spreads during the day.
§ The absolute, time weighted spread = STTR * tick-size
tick size / = 0,05 (e g VOLV B) / STTR / STTR / STTR / STTR# ticks / 1,00 / 1,40 / 1,50 / 2,30
Spread time on “# of ticks” =1 / 100% / 60% / 40% / 0%
Spread time on “# of ticks” =2 / 0% / 40% / 40% / 70%
Spread time on “# of ticks” =3 / 0% / 0% / 10% / 30%
Absolute time weighted spread / 0,05 / 0,07 / 0,075 / 0,115
· Spread Adjustment Factor - SAF (used in option 2 in ESMA’s consultation). If the SSTR is below 2, the tick size is considered to be constraining further reduction in the spread. In those cases a SAF is applied to reduce the tick size, meaning an adjustment of a certain stock’s tick size band relative to its natural state. Example: a stock traded in the 50-99,95 interval, will instead use the tick size for the 20-49.98 interval (e.g. in practice a reduction of the tick size with 40-50%).
Stock price / Tick size / Tick size with SAF-adjustment0,01
20-49.98 / 0,02 / 0,01
50-99,95 / 0,05 / 0,02
100-199,9 / 0,1 / 0,05
200-499,8 / 0,1
· Liquid market. The models use different definitions of “liquid market”.
- Option 1. If the stock has more than 500 trades per day, its considered as “liquid”
- Option 2. Use the standard definition of “liquid market” from MiFID-I.
Models
ESMA proposes to implement one of two alternative models referred to as option 1 and option 2
· Option 1. Two-dimensional (double entry) tick size table.
- The proposed tick size tables have been developed by using an algorithm to determine tick size that meets the requirement of a “spread to tick ratio” of 1,4-2,5 for liquid stocks and 1,4 for illiquid stocks.
- Model resulting in a matrix with two dimensions
§ liquidity profile
· “# trades per day” determines liquidity classification (>500 is “liquid”).
§ Price
· (three levels of increments (1,2 and 5), an increase from FESE-2 (previous: 1 and 5)
· Option 2. Single common tick size table (modelled broadly on existing FESE-2)
- The proposed tick size is based on a calculation of the Spread-to-tick-ratio
- A Spread Adjustment Factor is used to adjust the tick size to meet the requirements
- Determined by liquidity and price of the instrument (“liquid” or “non-liquid” according to the current MiFID-I definition).
- three levels of increments (1,2 and 5), an increase from FESE-2 (previous: 1 and 5)
Major differences compared to today’s model
The major differences between current implementation and the proposed new models are:
· One, single model for whole Europe, governed by ESMA
- “it is clearly the intent of MiFID II that the setting of tick sizes be under the control of the regulators rather than individual trading venue operators” (4.8 s 289 DP 2014)
· Price increments will be 1, 2 and 5 in both new models, compared with 1 and 5 in FESE-2, i.e. the tick size levels will be more granular.
· The tick size levels will be reduced in both new models, and for liquid stocks in particular.
Differences and similarities between the models
The similarities between the models:
· Both models are targeting a STTR of 1,.4-2,5 for liquid and 1,4-5 for illiquid stocks
· Both models rely strictly on metrics
The difference between the models:
· Definition of “liquid” is different between the models.
- Option 1: “instruments with number of trades exceeding 500 per day”
- Option 2: “according to definition in MiFID-I, Article 22 (MiFIR)”
· Option 1 use an algorithm to find the appropriate tick size in order to achieve an expected STTR target (ex ante). If the calculation has been correct, the STTR will hopefully end up between 1,4 and 2,0.
· Option 2 has FESE-2 as starting point, and adapts AFTER calculating the actual FTTR (ex post).
- If the stock has an FFTR of < 1,4 the tick size will be adjusted (using a “Spread Adjustment Factor) so that a smaller tick size will be used (in order to move the STTR above 1,4)
- If the stock has an FFTR of ≥ 2, as above, but I higher tick size will be used,
· Option 1 uses the definition of “liquid stocks” as “number of trades per day exceeding 500”. Option 2 uses the current definition as in MiFID-I: 1) Free float of 500MEUR 2) number of trades exceeding 500 OR 3) daily turnover of 2MEUR. This will be modified to the new definition in MiFID-2, in due course.
Effects
Effects on all stocks
An analysis of the stocks listed on NASDAQ OMX Nordic shows that a significant number of stocks will get a SAF adjustment resulting in decreased tick size. In the table below: number of stocks affected if STTR <2.0 (current ESMA proposal) and if STTR < 1,4 (a suggested more conservative approach)
Effects on the index stocks (option 2)
Having in mind that the target spread on the blue chips stocks should, according to ESMA, be around 6-7 bp, its interesting to do some calculations on the index stocks on NASDAQ OMX.
· Theo unw (Theoretical unweigthed). Theoretical spread if the STTR is 1.0 (always on tightest possible spread), using equal weight for all stocks.
· Theo wt (Theoretical weigthed). Theoretical spread if the STTR is 1.0 (always on tightest possible spread) using index weights for all stocks.
· Adj wt (Adjusted weighted). Time weighted spread, adjusted with STTR, with index weights. This value is close to the “true” spread for the index.
· SAF adj wt (SAF adjusted weighted index). Those stocks with STTR < 1,4 get a SAF-adjusted new tick size.
· # SAF adjusted stocks. Number of stocks that get a new tick size due to SAF adjustment. STTR value of < 1,4 and 2,0 has been used.
theo unw / theo wt / adj wt / SAF adj wt (1,4)/ s30 / 6,1 / 5,5 / 7,0 / 5,2
/ c20 / 5,2 / 4,9 / 8,2 / 7,4
/ h25 / 5,5 / 5,6 / 8,3 / 7,0
Two important observations is that
· Today´s “ average weighted spread” for OMXS30, OMXC20 and OMXH25 is not far from 6-7 bp.
· There is room for a more conservative adjustment than the ESMA proposal. If only stocks with STTR < 1,4 (rather than the proposed 2.0) get a tick-size reduction, it will result in a weighted spread of around 7 bp.
Conclusion and recommendations
The proposed model to let ESMA decide on regulatory technical standards, in order to get a harmonised tick size model, based on metrics is positive. Both option 1 and option 2 is usable options, with some adjustments from the proposal in ESMA’s consultation:
· Option 1. “Average Daily Turnover is a better proxy for liquidity than number of trades.
· Option 2. If option 2 is used, the limit for when the SAF is used should be 1,4 not 2.0. Implementing quantitative metrics for determining correct tick size is a big step. The consequences of implementing a model resulting in a radical decrease (or increase), can be detrimental for the quality of the market, therefore it is of utmost importance to set the parameters with caution.
Other comments:
· It’s crucial that the scales are accessible via an electronically readable manner.
· Changes in tick sizes should be communicated in advance
· When moving from the current regime to the new regime, there should be room for shorter evaluation periods, let’s say 6 weeks, in order to minimise calibration issues.
· There should be room for adapting special rules for some stocks if found necessary.
· “Number of trades” is not a god proxy for “liquidity”, it should be combined with “Average Daily Traded (measured in MEUR)”. Furthermore, manually reported trades should not be a part of the calculation.
Appendix A - Tick size tables
Find below, the new proposed tick size tables. Highlighted in yellow are the new price levels.
New proposed tick size table / Current model (NasdaqOMX)Option 1
Opt 2- Liquid / Opt 2- Illiquid / FESE-II
Most liquid / Other
-0,4999 / 0,0001 / 0,0005 / -,4999 / 0,0001 / -4,99 / 0,01
0,5-0,9995 / 0,0005 / 0,0002 / 0.5-0.9995 / 0,0005
1-1.999 / 0,001 / 0,002 / 1-1.999 / 0.001
2-4.999 / 0,002 / 0,005
5-9.995 / 0,005 / 0,001 / 5-9.995 / 0,005 / 5-14,95 / 0,05
10-19.99 / 0,01 / 0,02 / 10-49,90 / 0,01 / 15-49.90 / 0,1
20-49.98 / 0,02 / 0,05
50-99,95 / 0,05 / 0,1 / 50-99,95 / 0,05 / 50-149,75 / 0,25
100-199,9 / 0,1 / 0,2 / 100-499,9 / 0,1 / 150-499,50 / 0,5
200-499,8 / 0,2 / 0,5
500-999,5 / 0,5 / 1 / 500-999,5 / 0,5
1000-1999 / 1 / 2
2000-4998 / 2 / 5
5000-9995 / 5 / 10
10000-19990 / 10 / 20
Appendix B– Metrics
Calculations are done for instruments in the MiFID-database (this means that double listed stocks like AZN, TeliaSonera, StoraEnso etc only appears on the listing venue).
XSTO
XHEL
XCSE
Appendix C– Proxy for liquidity
There is no simple metric to assess liquidity in a stock.
When a portfolio manager wants to shift stocks in a portfolio, or when a retail client want to buy or sell a stock, the consideration (price * quantity) is more important than how many trades, the order might result in .
[1]The data and calibrations are kindly provided by SEB.