Liquidity of the Chinese Agricultural Futures Market and Its Impact on Futures Price—Based on High-Frequency Data
Round 1
Reviewer 1 Report
According to lines 209 to 211 two subsets of data are used in estimations. However it is very difficult to know which subset is explained in data description (line 240, Table 1). More explanations should be added.
In line 218 the subject is “I”, in line 540 the subject is “Paper” and in line 541 the subject is “we”. It is better to use only one subject instead of three different.
Tables 3 and 4 which are explained in the text are not presented. They should be included into the paper. The reader could not believe in explanation of the tables which do not exist. I think mistakenly these tables are removed, but in order to evaluate I have to see them.
Daily data usually have not unit root. However, it is important to prove the absence of unit root in time series in data description (Table 1).
Only R2 is not enough to confirm the robustness of the estimations. The authors have to conduct some additional post-estimation analysis and prove that the estimations are robust.
Author Response
Point 1: According to lines 209 to 211 two subsets of data are used in estimations. However it is very difficult to know which subset is explained in data description (line 240, Table 1). More explanations should be added.
Response 1: Thanks for your suggestion! We have supplemented data description in Table 1 (line 318, line 341), Figure 2 (line 353).
Point 2: In line 218 the subject is “I”, in line 540 the subject is “Paper” and in line 541 the subject is “we”. It is better to use only one subject instead of three different.
Response 2: Thanks for your suggestion! We have changed all personal pronouns to “we”, but still remain the subject “paper”. We will give reason as follows. On the one hand, the journal has no special requirements for that; on the other hand, we think that different subjects are suitable for different context.
Point 3: Tables 3 and 4 which are explained in the text are not presented. They should be included into the paper. The reader could not believe in explanation of the tables which do not exist. I think mistakenly these tables are removed, but in order to evaluate I have to see them.
Response 3: We are so sorry for carelessness. We have supplemented Tables 3 and 4 in revised paper.
Point 4: Daily data usually have not unit root. However, it is important to prove the absence of unit root in time series in data description (Table 1).
Response 4: Thanks for your suggestion! We have supplemented the results of unit root test in Table 1.
Point 5: Only R2 is not enough to confirm the robustness of the estimations. The authors have to conduct some additional post-estimation analysis and prove that the estimations are robust.
Response 5: Thanks for your suggestion! We divide the sample intervals into 4 subsample sections and do regress analysis within each subsample section, finding that there are no significant differences in the sign and significance of the estimated coefficients. So, we think our results are robust.
Soybean future | corn future | |||
PV | NV | PV | NV | |
Β0 | 0.005283(7.967) | -0.01533(-5.604) | 0.000321(6.598) | -0.000311(-7.567) |
Β1 | -0.000266(-1.211) | 0.000561(1.994) | 0.000078(0.997) | 0.000019(1.112) |
Β2 | -0.000189(-0.543) | 0.000343(1.252) | 0.000101(1.992) | -0.000076(-2.312) |
γ open,0 | -0.000232(-1.442) | -0.000795(-1.167) | 0.000045(0.923) | -0.000084(-2.768) |
γ open,1 | -0.000117(-0.627) | 0.000066(0.107) | -0.000023(-1.221) | 0.000032(0.755) |
γ close,0 | -0.000322(-1.172) | -0.000376(-0.606) | 0.000078(1.822) | -0.000052(-1.998) |
γ close,1 | -0.000723(-1.870) | 0.000887(1.3000) | -0.000230(-2.132) | 0.000045(2.011) |
Sum | 0.004828 | -0.000629 | 0.000500 | -0.000368 |
Lags | -0.000455 | 0.000904 | 0.000179 | -0.000057 |
Open_S | -0.000349 | -0.000729 | 0.000022 | -0.000052 |
Close_S | -0.001045 | 0.000511 | -0.000152 | -0.000007 |
A1 | 0.02691 (0.643) | -0.25866 (-4.850) | ||
A2 | 0.07870 (1.888) | -0.07096 (-1.333) | ||
C | 0.01548 (1.568) | -0.19330 (-0.128) | ||
Adj.R2 | 0.554 | 0.611 | ||
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The paper is potentially a valuable contribution in the field of the studies on the futures markets yet it requires considerable improvements. To begin with, it requires careful copyediting due to many errors or inconsistencies (e.g., Authors use interchangeably 'we' and 'I' throughout the manuscript). I present the remaining comments by referring to various sections of the paper:
Introduction:
- aim of the paper should be emphasized as it is difficult to spot; there are in fact two aims listed in the introduction and they are not enirely consistent; aim in the abstract is much more clear;
- Authors could list the main conclusions in this Section as suggested in the journal's editing requirements;
- literature review is mostly correct and up-to-date references were used; however, I suggest using some references about high-frequency trading as the title of the article implies that it is covered in the paper;
- Authors should explain how does the paper fit into the journal's main scopes, i.e. why would be of interest to its readers; in the current form it seems more fit for some financial journal.
Section 2 is generally correct but I recommend substantially extending Section 2.3. as it is of the utmost importance in the context of the presented study. I also think that adding some paragraphs about algorithmic trading on the Chinese futures market could enhance the paper as it may also influence the discussed issues.
Authors should present some justification of their choice of the analyzed time period, e.g., whether it is representative and may be used to formulate generalizations (maybe some unusual trading patterns occured over this period?). I also believe that conclusions about the U-shaped relationship may be too farreaching and they are not supported by the evidence on the figure, in particular with regard to the part (a) of the Figure 2 .
Section 4 should be significantly revised and in fact rewritten:
- its title is misleading - I suggest dividing it into two parts, one strictly methodological and second with the presentation of the empirical results;
- title of the Section 4.1. needs to be changed as it is too general (Section 4.2. also includes regression framework);
- analysis seems correct from the methodological perspective yet presentation of the results must be improved as it is difficult to follow, with too many long statements;
- most serious: Authors refer to tables 3 and 4 that were not included in the manuscript - it hinders the evaluation of the key empirical section of the paper therefore I omit this aspect of the review.
Explanations in the Section 5 do no take into account the possible impact of the automated trading that could also affect the analyzed relationships. Moreover, Authors must place their results in the broader context and compare them to the previous studies in this field, not necessarily about the Chinese market. They could thus underline their contribution as in the current manuscript it is unclear in what ways (if any) does the paper bring anything new to the financial studies on the futures markets and agriculture. Authors should address the limitations of the applied research methods.
Conlusions repeat to high extent information from the previous sections - it would be adviseable to focus more on the implications of the study for professionals and academics. Moreover, the paper could be concluded with the formulation of the directions of the future research.
Author Response
Point 1: The paper is potentially a valuable contribution in the field of the studies on the futures markets yet it requires considerable improvements. To begin with, it requires careful copyediting due to many errors or inconsistencies (e.g., Authors use interchangeably 'we' and 'I' throughout the manuscript).
Response 1: Thanks for your suggestion!
We have changed all personal pronouns to “we”, and polished it overall with the help of native speakers.
Point 2: Introduction:1)aim of the paper should be emphasized as it is difficult to spot; there are in fact two aims listed in the introduction and they are not entirely consistent; aim in the abstract is much more clear;2)Authors could list the main conclusions in this Section as suggested in the journal's editing requirements;3)literature review is mostly correct and up-to-date references were used; however, I suggest using some references about high-frequency trading as the title of the article implies that it is covered in the paper;4)Authors should explain how does the paper fit into the journal's main scopes, i.e. why would be of interest to its readers; in the current form it seems more fit for some financial journal.
Response 2: Thanks for your suggestion!
1) Based on measuring accurately liquidity index from high-frequency data, our objectives in this paper are to better understand the impact of intraday trades on futures price from information asymmetry perspectives, and to explore how daily liquidity affects asset pricing and its day-of-the-week effect.
2) We have supplemented main conclusions in “introduction” Section. Such as, line 71-74, line 91-99 and line 138-147.
3) Follow your suggestion, we add extra references about high-frequency trading as the title of the article. For example, Yacine and Yu (2009) and Seifoddin et al. (2017) used a multifactor asset pricing model and explored the extent to which this liquidity factor is priced when high frequency data are available. In the paper, we generalize them into one expression, namely “Empirically, different measurement methods or sample data may bring diverse results [20-25]” (line 131-132).
Reference: 1)Yacine Aït-Sahalia, Yu J. High Frequency Market Microstructure Noise Estimates and Liquidity Measures[J]. Annals of Applied Statistics, 2009, 3(1):422-457.2) Seifoddin J., Rahnamay R. F., Nikoomaram H. High Frequency Market Microstructure Noise Estimates and inference regarding returns: a portfolio switching approach[J]. Financial knowledge of security analysis, 2017 10 (34):1-12
4) In view of critical role futures markets play in guiding sustainable agricultural production, we think that a better understanding of liquidity costs and liquidity pricing is of great significance to a sustainable development of the agricultural commodity market in China. In other words, the answers for the problem to be solved in the paper contribute to providing guidance about futures trading efficiently and giving a further boost to the sustainability of Chinese agricultural futures market. So, the paper fits into the Sustainability journal's main scopes.
Point 3: Section 2 is generally correct but I recommend substantially extending Section 2.3. as it is of the utmost importance in the context of the presented study. I also think that adding some paragraphs about algorithmic trading on the Chinese futures market could enhance the paper as it may also influence the discussed issues.
Response 3: Thanks for your suggestion!
1) We have expanded Section 2.3 (line 252-275)
2) We agree that algorithmic trading is an important content to study roundly market liquidity. Meanwhile, algorithmic trading remains an essential ingredient to achieve best execution and reduce transaction costs. Liquidity traders, without superior information, could only estimate the value and liquidity of futures contract according to transaction records, and utilize the algorithmic trading to determine further their optimal order. According to the weaker connection between algorithmic trading and the question discussed in paper, we do not introduce algorithmic trading with very detailed notes, but we think it is the directions of the future research.
Point 4: Authors should present some justification of their choice of the analysed time period, e.g., whether it is representative and may be used to formulate generalizations (maybe some unusual trading patterns occurred over this period?). I also believe that conclusions about the U-shaped relationship may be too far-reaching and they are not supported by the evidence on the figure, in particular with regard to the part (a) of the Figure 2.
Response 4: Thanks for your suggestion!
1) In view of the availability of high-frequency data, we choose latest data sets from January, 2016 to April, 2018. We divide the sample intervals into several subsample sections and do regress analysis within each subsample section, finding that there are no significant differences in the sign and significance of the estimated coefficients. So, we think our sample data can produce robust results.
2) When separating day trading and night trading into individual sections, our results support previous findings about U-shaped pattern of intraday liquidity, that is that the liquidity is higher during the opening and closing periods than middle of trading hours. The curvilinear trend in Figure 2(a) look like “W”, because trading sections in the soybean market include day trading and night trading, which contain both the opening and the closing.
Point 5: Section 4 should be significantly revised and in fact rewritten:1)its title is misleading - I suggest dividing it into two parts, one strictly methodological and second with the presentation of the empirical results;2)title of the Section 4.1. needs to be changed as it is too general (Section 4.2. also includes regression framework);3)analysis seems correct from the methodological perspective yet presentation of the results must be improved as it is difficult to follow, with too many long statements;4)most serious: Authors refer to tables 3 and 4 that were not included in the manuscript - it hinders the evaluation of the key empirical section of the paper therefore I omit this aspect of the review.
Response 5: Thanks for your suggestion!
1) Following your suggestion, we have readjusted the structure of this section, dividing it into two parts: 4.1 model introduction (including 4.1.1 “The regression framework of price movement on trading volume” and 4.1.2 “A modified model of Asset pricing”) and 4.2 result analysis (including 4.2.1 “The impact of trading activity on futures price and its intraday seasonality” and 4.2.2 “The role of liquidity in asset pricing and its weekday seasonality”)
2) We change the title of 4.1.1 section to “The regression framework of price movement on trading volume”
3) We have modified complex expression and reduce long sentences.
4) We are so sorry for carelessness. We have supplemented Tables 3 and 4 in revised paper.
Point 6: Explanations in the Section 5 do not take into account the possible impact of the automated trading that could also affect the analysed relationships. Moreover, Authors must place their results in the broader context and compare them to the previous studies in this field, not necessarily about the Chinese market. They could thus underline their contribution as in the current manuscript it is unclear in what ways (if any) does the paper bring anything new to the financial studies on the futures markets and agriculture.
Response 6: Thanks for your suggestion!
1) We agree that automated trading could also affect the analyzed relationships. Sophisticated traders usually base algorithmic trading strategy on the market trends which are determined by using statistics, to make profits. In fast-moving markets, getting in or out of a trade a few seconds earlier can make a big difference in the trade's outcome (line 690-693). Therefore, automated systems are increasingly applied to monitor the markets for trading opportunities and execute the trades as soon as trade criteria are met, since computers respond immediately to changing market conditions (219-222). Obviously, this would contribute to forming certain special trading situation, a wait-and-see state or frequent trading condition.
2) Following your suggestion, we extend the comparison with previous studies (line 715-719) and contribution (line 653-657, line 686-688)
Point 7: Conclusions repeat to high extent information from the previous sections - it would be advisable to focus more on the implications of the study for professionals and academics. Moreover, the paper could be concluded with the formulation of the directions of the future research.
Response 7: Thanks for your suggestion!
1) Following your suggestion, we reduce the repetitive information in “conclusion” section.
2) As for the implication of the study for professionals and academics, we conclude that it encourages researchers to take heterogeneity among markets into account and to focus on the different features of price impact among varieties (line 742-743).
3) We also point out the limitation of this paper (line 743-745) and directions of the future research (745-747).
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The paper is improved enough for publication in an academic journal.
Reviewer 2 Report
The paper has been revised correctly.
