Abstract
The purpose of this article is to study accounting valuation practices in French bankruptcies following enactment of the Law of 1838. The research is based on a study of 500 files in the archives of the Paris Court of Commerce. After first presenting the main steps in the bankruptcy proceedings, it is shown how the self-interest of the agents – the bankrupt, the receiver, and the creditors – was built up through accounting valuation practices. The outcome of this analysis provides an insight into accounting valuation practices in nineteenth-century France and a better understanding of the self-interest of the agents in bankruptcy proceedings.
Bankruptcy proceedings are very rich sources for accounting historians. Besides the fact that they have been very well preserved, these archives confirm the importance of the law in the history of accounting practices (Napier, 1998). Until now, few works have used such archives. Most Anglo-Saxon research has focused on the relationships between accountants and lawyers who “are competitive but may also be co-operative” (Walker, 2004a: 247). This is the result, in part, of national characteristics and the emergence, from the nineteenth century, of a structured accounting profession in Great Britain, whereas this phenomenon appeared later in France (Ramirez, 2001). Moreover, the history of continental European accounting is quite different. Colbert’s Ordonnance of 1673 influenced many corporate and bankruptcy laws in Europe (Walton, 1993).
The history of the links between accounting and law can be studied from several points of view: the history of law (Edey and Panitpakdi, 1956), the history of audit practices (Maltby, 1998), the history of the professions (Walker, 2004a), the history of justice (McKinstry et al., 2002), or the history of bankruptcy. The last aspect seems to be little studied today. Literature on the history of bankruptcies exists, but relates to other disciplines such as the law, economics and sociology. Law historians ask questions about the conception of the laws on the one hand and their judicial impact on the other (Duffy, 1985). Bankruptcy also concerns economists, if only through the links between many bankruptcies and economic cycles (Moss and Hume, 1983). Bankruptcies are also a fertile source of information on sectoral and regional developments (Hoppit, 1987; Solar and Lyons, 2011). Finally, bankruptcies also concern wide swathes of social history and can be used as sources (Lester, 1995) in elaborating social history. There are few studies about many European countries, although the history of bankruptcy law and its efficiency in Italy (Di Martino, 2005, 2008), Germany or Sweden (Gratzer, 2008) offers some insights about differences with Great Britain and United States.
In the French case, the same type of distinction is evident. For legal historians, the genesis of laws (Choffée, 1997) and their application is examined (Noël, 2007). For economic historians, the questions raised concern the effectiveness of judicial proceedings (Hautcoeur and Levratto, 2006) and the long-term historic development of these laws (Marco, 1985). For certain commentators, bankruptcies are, therefore, a source for the social history of bourgeois society (Daumard, 1963), shopkeepers (Coquery and Praquin, 2008), and the Chambers of Commerce (Lemercier, 2008). For management historians, it is possible to understand the causes of bankruptcies (Marco, 1981; Malécot, 1981, 1985). Finally, some accounting historians have used bankruptcy archives to study the diffusion of accounting in France during the eighteenth (Lemarchand, 1994) and nineteenth centuries (Labardin, 2011).
French and English bankruptcy law, historically, were quite different: whereas since the beginning of the eighteenth century English law had encouraged the bankrupt “to co-operate with the recovery of his assets for distribution to his creditors” (Lester, 1995: 38), French law sought to punish the bankrupt (Choffée, 1997). Nonetheless, in France, as in Great Britain, prison sentences were possible. The French specificity is the consequence of the belated structuring of accountancy (Bocqueraz, 2001; Ramirez, 2001).
The anticipated contribution of this type of research is three-fold. Firstly, this research is helpful in describing the accounting practices of French small and medium sized firms in the nineteenth century. There are no official statistics in nineteenth-century France about the number of businesses; however, according to the fiscal statistics of French business licence (patente), there were 1,443,778 businesses in 1847 and 1,966,493 in 1887 (Jobert, 1991: 35). Moreover, French small and medium sized businesses represented an important part of the economy in the late nineteenth century (Piore and Sabel, 1984). As such, the accounting valuation practices of small and medium sized firms are a major issue.
Nevertheless, most French research on accounting history has concerned major firms. Since Lemarchand’s (1993b) work, some historical information on the accounting valuation practices of companies in France has been available. For current assets, market value was used, but for capital assets, the matter seems far less clear (1993b: 507–511): market value and historic cost were used without either achieving pre-eminence. Nevertheless, the prudence concept was commonly accepted during the nineteenth century in France (Lemarchand, 1993b: 505–507) and in England (Maltby, 2000).
Secondly, this research allows us to elucidate an economic and legal issue – business failure and bankruptcy. More concretely, it helps us to understand the mechanisms of the proceedings by focusing notably on accounting issues. Whilst other studies on the same type of theme centre on the long-term and the impacts of professionalisation and involvement in judicial proceedings (Coquery and Praquin, 2008), this research places each of the accounting valuations in its own context to highlight the self-interest of each of the agents.
Thirdly, such research opens up these relatively technical archives to subsequent researchers to explore in greater depth. Its true worth lies in providing us with a better understanding of how bankruptcy proceedings work, enabling us to apprehend what motivates the agents involved and therefore the meaning of the positions they take. In this respect, the role played by accounting proves to be of particular significance. The purpose of this article is to bring to light the role of accounting valuation in bankruptcy proceedings, thereby providing an idea of the accounting self-interest of the various agents involved.
The remainder of this article is divided into five sections. The first defines more precisely the interplay between accounting valuation and self-interest. In the second we detail and justify the methodology followed. The third recalls the legal framework for bankruptcy in France with respect to the main agents involved in the proceedings and briefly refers to the legal texts governing it. This section aids in defining the relationships between the agents and reveals their interests. The penultimate section compares the various balance sheets disclosed in order to extrapolate the behaviour that were engaged in by the agents in the bankruptcy process. The concluding section provides a clearer idea of the links between accounting valuation and self-interest by the three main agents involved in this process.
The interplay between self-interest and accounting valuation in bankruptcies
In the case of the bankruptcies studied here, accounting plays a dominant role. It serves, in fact, to distribute the assets between the various stakeholders. The work of Coquery and Praquin (2008) placed the accent on practical valuations adopted in connection with the legalisation and professionalisation of the proceedings. Basing this study on the idea of self-interest (Roberts and Jones, 2009) highlights the construction of practices of accounting valuation in French bankruptcy cases of the nineteenth century.
Whilst the classic conception of the economy puts forward the idea of autonomous, isolated agents, Callon (1998), on the contrary, puts forward the idea of agents who are an integral part of their environment and networks. It is from these conditions that the self-interest of the agents can be built up (Roberts and Jones, 2009: 857). It is only later that the market proves a means of coordination that enables interests that are at the same time divergent and overlapping to be reconciled.
The application of this approach proves particularly interesting here since the agents are locked in judicial proceedings – bankruptcy – which determine the times and conditions under which accounting comes into play. With this context as a starting point, an analysis of accounting balance sheets uncovers recurring practices. This highlights the links between the self-interest of the various players and accounting valuation practices. The approach adopted demonstrates the place of self-interest both in the judicial context and in the players’ networks. Accounting plays a particular role here: calculating tools “do not merely record a reality independent to themselves: they contribute powerfully to shaping, simply by measuring it, the reality they measure” (Callon, 1998: 23). In our case, accountancy fulfils these two functions: each player obviously seeks to express reality, but also, via accounting valuation, to defend his own interests.
Choosing this theoretical anchor brings to light new dimensions. For example, contrary to Coquery and Praquin (2008: 67) who explain the accounting evaluation of mistakes by simple habit, this study shows how this practice can form part of the construction of self-interest. This shows that even in practices that are apparently outside the market, such as bankruptcy, the calculation tools contribute to creating reality.
Methodology
An accounting historian wishing to work on accounting practices in French bankruptcies has few sources. The accounts of bankrupts were not preserved, unless they chose to leave them to archives (which was very rare). In cases of bankruptcy, the local press published the call for creditors at the beginning of the proceedings; however, relying on this source material raises several problems. First of all, it would be difficult to constitute a homogeneous sample because the information is so widely scattered. Secondly, there are no accounting details in such notices. Consequently, this research is based on bankruptcy archives, which allowed comparisons of between one and three balance sheets for each bankruptcy.
Poor preservation of bankruptcy files stored in the archives of many French regions 1 led to a focus on those held in the Paris region, which possess several tens of thousands of files dating back to the nineteenth century. Moreover, statistics show that roughly a quarter of all bankruptcies during the time period under investigation occurred in this region (Jobert, 1991: 151, 175–176).
To detect possible changes in practices, 100 bankruptcy files were systematically consulted every 10 years, making a total of 500 (see primary sources). These files are organised chronologically. The purpose of the 10-year gap was to compare any shifts over the 40 years that are the focus of the study. This method of gathering data resulted in a focus on small bankruptcies – the overwhelming majority of all cases – including a significant number of small traders and artisans who did not operate as companies. This method prevented aggregation of the five samples, because the number of bankruptcies differed from year to year.
It is also worth providing some evidence to support the methodological approach. Working on the first 100 files allowed study of the work carried out by receivers. The number of different receivers nominated by judges over a given period of time could be studied, from which could be deduced the average amount of time each receiver spent on a single file.
A non-random process raises the question of its representativeness. However, it is difficult to imagine why a January bankrupt would have kept his books better than a June or September bankrupt. The theoretical average number of bankruptcies was compared with the average number of the sample: four out of five observed averages are close or very close to the theoretical average. There is only a little difference for the year 1867. This provides further evidence that the sample is representative (see Table 1).
Analysis of the representativeness of the sample (1847–1887).
Note: a So as not to falsify the analysis of the time period, incomplete and missing files are included in the analysis.
In any case, the size of the sample is similar to, if not larger than, samples provided in other research published to date (Coquery and Praquin, 2008; Hautcoeur and Levratto, 2006). Moreover, selecting the first 100 files of each year allows comparison of the same period in each year, thereby neutralising seasonal factors.
The sample studied includes all kinds of commercial bankruptcy, where the aim to understand the mechanisms behind all types of bankruptcy implying examination of both small and large bankruptcies. This gives a better understanding of accounting practices generally, rather than – as is often the case – favouring large companies.
Marco (1997: 118–119) offers some insights about the sectoral evolution of bankruptcies in the whole of France (1874–1896) 2 (see Table 2).
Number of bankruptcies in France (1874–1887).
A more detailed sectoral analysis provides further information on this evolution: the number of bank and textile bankruptcies diminished, but other businesses such as inns, bars and building contractors increased (Marco, 1997: 117–120). Several causes may explain such evolution: the concentration of some sectors entailed a rather limited number of firms, the importance of the economic cycle hit more some sectors than other, etc.
For the purposes of this research, only files that were significantly incomplete were excluded, that is, cases where no inventory is provided or the outcome of bankruptcy proceedings is unknown. In itself, the very fact of taking stock of the number of bankruptcy proceedings provides a better understanding of shifts in practices over this period. A few rare cases were retained where the outcome is unknown but bankruptcy proceedings had begun with the carrying out of an inventory. The end result was the exclusion of only a very few files from the sample.
To assess the size of a business, liabilities are probably the most appropriate measure (see Table 3). Firstly, uncertainty surrounding the value of assets makes choosing liabilities preferable. Secondly, a number of bankrupts would deplete all their resources before filing for bankruptcy, thereby presenting virtually no assets when, in fact, their business was actually quite sizeable.
Share of liabilities for the bankruptcies studied (1847–1887) a .
Note: a The amount of liabilities the bankrupt declared is used. From 1867 onwards, the number of bankrupts who did not file a balance sheet rose considerably.
In each file the same data was systematically reviewed: name, profession and residence of the bankrupt, date on which bankruptcy was filed, name of the receiver(s), amount of assets and liabilities at the beginning of the bankruptcy proceedings, its outcome, any legal proceedings stemming from the bankruptcy, the receiver’s comments on the quality of bookkeeping and, finally, the bankrupt’s bookkeeping as described in the inventory.
The use of this data has been two-fold. Whenever possible, that is to say when the information was recurrent (amount of liabilities, issuing from the proceedings, etc.), this has been incorporated into statistical tables. On the other hand, certain information is less frequent (commentaries on the guiding principles of accounting valuation on the habits of bankrupts, etc.). When such information seems to explain observations made elsewhere, this has been integrated into the analysis.
The construction of the agents’ self-interest
The rules of the game: Bankruptcy proceedings according to the Law of 1838
Ever since Colbert’s Ordonnance of 1673, bankruptcy in France had been governed by written law. Indeed, the 1807 Code de Commerce reiterated several dispositions of the Ordonnance. In the name of improving morality in trade and commerce, it was a matter of punishing the bankrupt who was suspected of wanting to extricate himself unduly from his obligation to pay his creditors (Choffée, 1997). To some extent, the Law of 1838 loosened the sanctions tied to bankruptcy in order to make the proceedings more efficient (Choffée, 1997: 157–177).
The bankruptcy proceedings may be summarised as follows: once a trader’s inability to pay his creditors became apparent, the bankrupt had three days to file for voluntary bankruptcy. If he failed to do so, his creditors or a magistrate could then take the initiative to declare him bankrupt. An initial balance sheet was then drawn up by the bankrupt (article 439). To facilitate comparison, 3 it is this balance sheet that is used here to estimate assets and liabilities. The magistrate then appointed a provisional receiver to take charge of carrying out proceedings on a daily basis. The creditors were then summoned by the magistrate within a fortnight to appoint a permanent receiver (article 462).
Several scenarios could subsequently arise. If, from the outset, assets were too low to cover the cost of the receiver’s fees and no creditor was willing to put up more funds, the receiver generally did not undertake any further proceedings and the bankruptcy proceedings were halted due to insufficient assets.
In most cases, the receiver promptly drafted an inventory of the bankrupt’s possessions. After preparing this inventory, the receiver drew up an initial report to present creditors on the state of the bankruptcy proceedings. He then presented a second accounting balance sheet of the bankrupt’s operations.
When creditors opted for a settlement, they relinquished their credits in exchange either for the bankrupt’s agreement to pay back a percentage of his debts from these assets (a simple settlement) or for an arrangement whereby they ceded their assets in full to the bankrupt who then committed to paying back some of his debts (a settlement with asset waiver). Any settlement included the scheduling of repayments was set by mutual agreement between the bankrupt and his creditors.
In the event of liquidation, the receiver provisionally administered bankruptcy proceedings in lieu of the bankrupt. Most often, the receiver confined himself to liquidating the bankrupt’s possessions and sharing out the sums raised among the creditors after his own expenses (as well as his fees) had been deducted. In the event of liquidation, and by virtue of article 538, the receiver was also obliged to raise the issue of the bankrupt’s legal liability (see Figure 1).
Bankruptcy proceedings.
Examination of the files enables measurement of statistical shifts in the use of the various proceedings (see Table 4). Two points emerge from these figures: first, the gradual decline in the number of settlements over the period as a whole; and, second, the growth in the number of proceedings halted due to insufficient assets. More and more files were closed for this reason and this phenomenon will be explained in the fifth part of this article.
The outcome of bankruptcy proceedings in the Paris region (1847–1887).
The impact of rules on the construction of the agents’ self-interest
To understand the mechanism behind bankruptcy proceedings, the various agents involved in the proceedings and their respective interests are detailed here. This will allow their accounting-related self-interests to be more fully grasped. Existing research and Acts are used to highlight the main motivations of the three principal agents in proceedings (bankrupt, creditors and receiver).
Naturally, the bankrupt is the central character. If he does not conceal a part of his assets, he has much to lose in these proceedings. Since organisations in the form of limited liability companies rarely figured in bankruptcy proceedings (Lefebvre-Teillard, 1985), most bankrupts risked forfeiture of their personal possessions. When businesses were closed down due to insufficient assets, bankrupts lost what little they may have possessed in all cases. Any lack of cooperation would also lead the receiver to suspect the bankrupt of not disclosing all his assets, thereby triggering legal proceedings for banqueroute frauduleuse. 4 In the more common case where there were sufficient assets to cover the cost of the receiver’s fees, the bankrupt had every interest in being conciliatory towards his creditors who would decide whether or not to grant him the benefit of a settlement. A settlement would allow him to avoid the seizure and sale of his possessions, that is, liquidation, in exchange for his promise to repay his creditors in part or in full.
For the creditors to underwrite such an agreement they still had to believe that the bankrupt was in a position to honour his commitments. The bankrupt therefore had to make a pledge of good faith (Hautcoeur and Levratto, 2006: 25). For ordinary creditors (i.e. those not preferred in asset distribution), bankruptcy proceedings entailed the near certainty of making a loss. The purpose of the proceedings was to limit such losses.
When there were sufficient assets, creditors had a choice between settlement and liquidation. The receiver’s first report helped them to reach their decision by providing important indications (circumstances surrounding the bankruptcy, estimates of assets, the bankrupt’s previous history, etc.). Following this initial report, creditors then voted on whether or not to reach a settlement. Article 507 stipulated that, “this settlement can only be reached with the agreement of a number of creditors forming the majority and, in addition, representing three-quarters of total credits as verified, disclosed or provisioned for”.
Here, the role played by the receiver was crucial. Through his inventory and then his report (in the case of settlement) or his two reports (in the case of liquidation), his expertise helped to inform the creditors’ decision. As to the identity of the bankruptcy receivers, Coquery and Praquin (2008: 64), present them as “jurists (with law degrees) or lawyers”. This research concurs with this hypothesis: in 1867, for instance, one receiver introduces himself as a doctor of law (file 7336), another as a lawyer (file 7402), and a third as a graduate in law (file 7390).
Other facts corroborate this hypothesis. During the Dalet bankruptcy in 1847, the receiver mentioned in his report that “an expert in bookkeeping has been appointed” (file 7281), indicating that he had no specialist expertise in accounting, contrary to the English case (Walker, 2004b). Only later did the formation of a body of chartered accountants working for the Paris Court of Commerce occur, as the work of Chauvaud (2002: 102) confirms: a category for “bookkeeping” only appeared in 1885. Such positions subsequently became highly prestigious for accountants because, in 1906, members of the Société Académique de Comptabilité – one of the first professional bodies in accounting in France – did not hesitate to present themselves as experts working for the Paris Court of Commerce. The receiver’s interests are more difficult to discern, although some insights will be provided in the fifth section.
The magistrates at the Court of Commerce warrant special mention. As Claire Lemercier (2008) reminds us, these were merchants, bankers or industrialists who were elected to office by their peers. Several factors suggest that the magistrate’s role was therefore low-key. In the cases studied here, magistrates generally confined themselves to concurring with receivers’ opinions. An explanation for this might be the gap between the low number of magistrates (20 according to Lemercier (2008: 66)) and the colossal number of court cases (70,000 by the end of the Second Empire according to Lemercier (2008: 66)).
An analysis of accounting valuation in balance sheets
A historian of accounting may find up to three balance sheets in a bankruptcy file. For simplicity’s sake, in the remainder of this article these three balance sheets will be called the bankrupt’s balance sheet, the settlement balance sheet, and the liquidation balance sheet, respectively.
This section gives some insights into accounting practices and accounting valuation. A comparison of accounting valuation in the three balance sheets (see Table 5) permits understanding of the construction of the self-interest of each agent and this will be presented in the fourth section.
Principles of asset valuation in bankruptcy (1847–1887).
The bankrupt’s balance sheet
When compared to modern accounting practices, the findings of the examination of bankrupts’ balance sheets are fairly unusual. Assets and liabilities were balanced in only a few exceptional cases. Out of the 100 files dating from 1847, only the Benistant (file 7277) and Riban (file 7291) bankruptcies provide a balanced balance sheet, and such cases remain few and far between for the other years – 8 in 1857 (files 13659, 13673, 13676, 13688, 13699, 13723, 13742 and 13759), 1 in 1867 (file 7408), 1 in 1877(file 3270) and 2 in 1887 (files 1555 and 1593). This means that potential losses were stated as assets and potential gains as liabilities.
Closer scrutiny of assets and liabilities enables better understanding of how a business’s patrimony was represented in accounting terms. Coquery and Praquin (2008: 59) remark that, in both the eighteenth and nineteenth centuries, “creditors and debtors were jotted down in a cursory style or in the form of a list”. In the files examined, the idea of capital contributions had not yet become widely accepted and understood. 5 In the large majority of cases (and notably for small liabilities), only debts were written down. This may be explained, firstly, by the impossibility of registering capital that, initially, was merely the difference between assets and debts, and secondly, by the meagre capital contributions that most bankrupts could muster. Furthermore, such shortages of capital were often mentioned as being one of the main causes of bankruptcy.
Valuing liabilities, on the other hand, posed few difficulties, since debts were registered at their historic cost. With respect to liabilities, the bankrupt could commit two types of fraud.
Firstly, he could seek to add bogus credits so that others close to him could then recover a part of the overall bankruptcy payout. A case in point was Folliot-Lenoir (file 7271), a bankrupt who, in 1847, was prosecuted for having overstated his credits. Any such accusation led to prosecution for fraud before the criminal courts. Generally speaking, receivers tended to discount any of the bankrupt’s relatives who presented themselves as creditors. In 1867, the receiver Beaufour attentively examined the funds put up by the bankrupt’s father before concluding in his report that, “the bankrupt’s father who had made a proposal to contribute capital of 77,187.45 francs does indeed appear to be a genuine creditor” (file 7330).
The second fraud that could be committed consisted quite simply of forgetting some debts. Any publicity surrounding the bankruptcy ruling made this a risky move: if a creditor who was omitted from the balance sheet came forward, the bankrupt ran the risk of arousing the suspicions of the receiver and the creditors who could then reject any settlement deal.
However, most problems arose from assets disclosed in the balance sheet. To illustrate how bankrupts went about valuing their assets, the best example is provided by Antoine Rebière (file 7236) 6 whose business went under in 1847. His assets, amounting to 73,898.77 francs, covered almost all of his liabilities, amounting to 77,713.39 francs.
It is difficult to know if asset disclosure followed any clear rules. Although valuations of cash register receipts were rarely debated in subsequent proceedings, this was not the case for other items: intangible assets were noted for memory 7 by the receiver in his valuation who considered that these were worth next to nothing (whereas the bankrupt had estimated their worth at 15,690 francs). This was merely the opinion of the bankrupt who saw in his past investments (rightly or wrongly) profit potential. It is clear that “the bankrupt’s valuation was carried out in the hope of keeping his business as a going concern” (Coquery and Praquin, 2008: 67).
The value of debts was registered at their face value with no account taken for any subsequent losses. For instance, the receiver estimated the value of Mr Rebière’s debts as 2,227.60 francs – only 10 per cent of their face value. Such differences can be explained by the fact that most credits registered as assets at the time of bankruptcy were hard to recover (if this were not the case they would already have been recovered). We can assume that, had they been easy to recover, bankrupts would have recovered these funds to avoid defaulting on their payments.
Bankrupts’ lack of understanding (or plain fraud) can also be spotted in certain omissions. For instance, in 1867 the entrepreneur behind the Legros construction firm forgot to disclose as an asset “a small property worth roughly 6,000 francs that he owns in Bourgneuf (but) which is subject to taxes of a more or less equal value” (file 7332).
The settlement balance sheet
After drawing up an inventory, questioning the bankrupt, and consulting the latter’s accounting, the receiver drafted a report with a view to a potential settlement. The purpose of this report was to explain to creditors the circumstances surrounding the bankruptcy and to give them the elements they needed to make a fully informed decision. In general, this report was at most four pages long and contained the following sections: presentation of the circumstances surrounding the bankruptcy and its causes; the drafting of an accounting balance sheet enabling creditors to determine what debts they could hope to recover; and finally, remarks pertaining to any possible fraud or negligence committed by the bankrupt. The goal of the balance sheet was neither to account for assets nor to give their liquidation value. It was simply a question of estimating the value the bankrupt could draw from them in the event that his debts were reduced and rescheduled.
One might well wonder how the receiver proceeded with the drafting of this balance sheet. To reconstitute the debts he had three means available: the bankrupt’s books, the bankrupt’s declarations and, finally, declarations (and recognition of debts) made by the creditors. Assets were valued on the basis of the inventory the receiver drafted immediately following the bankruptcy filing.
Inventories carried out by receivers took the form of bound papers often containing countless lists of all known assets, including assets relating to business operations (plant, equipment, raw materials, etc.), debts, sums available in cash, and also the bankrupt’s personal possessions. Although there were several different forms of company structure, the only structure that legally limited liability to capital contributions was the Société Anonyme 8 – the limited liability company (Lefebvre-Teillard, 1985). In other words, the bankrupt’s assets generally included his personal possessions and even those of his household. Accordingly, tables and chairs are listed in bankruptcy inventories. In the event of liquidation, the bankrupt could forfeit all of his possessions since these could all be sold off. 9 As an example of what may be learnt by comparing assets with the bankrupt’s balance sheet, the balance sheet of the bankrupt Rebière in 1847 (file 7236) is worthy of consideration.
Here, the difference between the two balance sheets is considerable: whereas the bankrupt estimated his assets at almost 74,000 francs, the receiver only valued them at 11,000 francs. This example is striking and compels investigation of reasons explaining the difference between these two valuations. 10 The settlement balance sheet took into account a certain number of expenses: for instance, the receiver had to keep business operations going in order to ensure the sale of goods and the payment of workers’ wages, rents, etc. – all costs that cut into the amount of assets. Therefore, the first difference between the two balance sheets stems from the fact that they were not carried out at the same time and that some operations had been performed or begun in the interim period. The more assets there were to recover (stocked goods to sell off, accounts receivable to recover, etc.), the greater the expense the receiver incurred in carrying out this task. It is, therefore, hard to imagine the assets listed by the receiver being of greater value than those declared by the bankrupt.
Another explanation for this variance between estimates relates to the nature of the assets accounted for in the valuation. Intangible assets (patents, fonds de commerce) were often noted down “for memory” and most receivers considered them worthless. This brings to light two different conceptions of valuation. For the bankrupt, the value of these assets mostly stemmed from the fact that he valued these assets at their acquisition price (that is to say, the historic cost). He did not necessarily perceive the fonds de commerce or any patents as costs, but rather, as investments and as a gamble on the future. The receiver’s reasoning, however, was somewhat different: he produced an estimate that was necessarily prudent. There was overvaluation of some assets in the bankrupt’s balance sheet: for example, the bankrupt’s estimate of his personal assets at 4,500 francs was revalued by the receiver at only 1,043 francs.
Other factors may have also played a part in influencing the outcomes of valuation. A lack of understanding of accounting on the part of most bankrupts resulted in their committing errors. Studying the bankruptcy files invites a questioning of the extent to which incompetence or fraud is evident. Such is the case with the steel and metal trader Rouet in 1867. The report notes that:
I am unable to register the sum of 2312 francs 80 centimes in the balance sheet as cash register receipts because I have found no such money at the time of inventory; if Mr Rouet disclosed them in his balance sheet, this was done by mistake due to the fact that this balance sheet was drawn up on the basis of elements in an older statement of his financial position. (file 7326)
This instance illustrates how important a trader’s bookkeeping was for the receiver. It also demonstrates how a lack of sound bookkeeping could complicate the receiver’s task.
This stark example leaves unanswered the question of whether all bankrupts were notoriously incompetent or if some of them perhaps manipulated their supposed incompetence. It is difficult to offer a general answer to this question and receivers themselves were rarely categorical in their assessments. In contrast, in cases where no errors were found, receivers did not hesitate to comment favourably in their report findings. In the Roussel bankruptcy in 1847, the receiver accordingly remarked: “Not a single act in Mr Roussel’s business career suggests any breach of faith on his part” (file 7222).
In a few marginal cases, a further reason may explain variance between valuations: an event could occur between the bankrupt’s and the receiver’s valuations that recapitalised the bankrupt. For instance, in the Rouchier bankruptcy in 1867 (file 7329), the taking into account of the son’s inheritance resulted in a doubling of assets (see also the Beauvais bankruptcy in 1887, file 1524).
One final reason may also explain such differences as to how possessions were valued. Of course, persons carrying out the valuations were not one and the same. However, more significant was the change in the purpose of the valuation itself. On the one hand, the bankrupt made his valuation with a view to keeping his business as a going concern: the primary goal of his accounting was to account for past investments. On the other, the receiver drew up a balance sheet with another goal in mind: his purpose was to provide the bankrupt and the creditors with a reasonable basis on which to negotiate a settlement. The lower the valuation was, the lower the baseline for negotiations between creditors and bankrupts. Indeed, in most cases of settlement, discounts granted to bankrupts were not as large as those suggested by receivers. By underestimating the value of bankrupts’ possessions, receivers were perhaps anticipating the overstated demands of creditors.
The liquidation balance sheet
If the creditors rejected a settlement (or if it could not be granted due to conviction for banqueroute frauduleuse), the next step was to move to liquidation. In this case, the receiver 11 then had to draft a third balance sheet to enable distribution of the value of the bankrupt’s assets among the creditors. There were fewer items on this balance sheet than on previous ones because a portion of the assets had already been sold off and the real goal at this point was to determine the dividend 12 creditors would be paid.
Consider the example of Pierre Moria, a tallow manufacturer who went bankrupt in 1857 (file 13722). He declared assets of 3,135 francs and liabilities of 8,510.35 francs. When drafting the settlement balance sheet, the receiver remarked:
As for the assets that the bankrupt declares to be worth 3135 f, they are far from being so large today. They are made up of his plant and equipment and have been estimated in the inventory to be worth 990 f. However, were he to be granted the favour of a settlement, these assets would have greater value in his hands, in which case their value can rise to 2000 f.
Evidencing the significance of these estimates, if there had been a settlement, the receiver estimated that these assets could have been worth 2,000 francs. By avoiding legal adjudication, the price obtained would probably have been higher. However, since creditors rejected a settlement, they would have to share between them a significantly smaller sum, namely 383.57 francs. It should also be noted here that in this bankruptcy, the receiver’s expenses were high – amounting to 2,210.45 francs – payment of which was made a priority.
The differences between the valuations carried out by the same agent require explanation. The first explanation relates to the timing of the valuation: between settlement and liquidation, costs were incurred for provisional management and asset sell-offs. Given that the receiver’s management costs were higher in the event of liquidation, assets in the liquidation balance sheet were consequently lower than those in the settlement balance sheet. The Moria bankruptcy does not appear to be an isolated case: during the Fletcher bankruptcy in 1857, the dividend announced by the receiver was 22 per cent in the event of settlement, but only 5–7 per cent in the event of liquidation (file 13751). A similar situation occurred in the bankruptcy of the Charfe and Cie general partnership in 1867 (file 7311): the balance sheet drafted at the outset foresaw a dividend of 35 per cent. However, the receiver’s report accompanying the settlement balance sheet only foresaw a dividend of 7 per cent in the event of settlement, and 5 per cent in the event of liquidation.
Yet, more importantly, the conditions for valuation were different. In cases of settlement, it was a matter of putting forward a valuation that enabled all parties to envisage keeping the business as a going concern. In the context of liquidation, it was a matter of selling off assets often at a cut price. In 1847, the receiver Heurtey stated with clarity: “Yet I must say that valuation of the equipment stated in the inventory was done with a view to continuing business and those legal proceedings would undoubtedly reduce the value of assets” (file 7221).
Ten years later, another receiver, Battarel (file 13732), said the same thing in his settlement report. Due to the conditions of liquidation, assets are mostly sold off at lower prices (file 7346). These observations lead to the basic characteristics of balance sheet valuations being deduced, as shown in Table 5.
Table 5 should be interpreted for what it is: it is not designed to convey principles explicitly stated by the agents involved. On the contrary, it simply conveys their practices. Since French accounting literature of the nineteenth century was unclear and uncertain in identifying the guiding principles of accounting valuation (Lemarchand, 1993b), it would be absurd to see anything other than a rationalisation of behaviours after the fact. The depreciation of assets mentioned here is worth explaining. This depreciation simply aims to observe the loss of the value of a property compared to its historic cost (that is to say, its purchase price). The receiver clearly errs on the side of caution, which leads him to reduce the value of the asset in the name of a loss linked to wear and tear over time.
The accounting agents’ self-interest
Having detailed valuation practices in bankruptcy proceedings, it is now possible to understand the ways in which agents appropriated the legal process in an attempt to exert influence over proceedings in favour of their own self-interests.
The accounting self-interest of the bankrupt
It is difficult to carry out a global analysis of the relationship of bankrupts to their creditors and receivers. One the one hand, each case is too specific to enable any global conclusion to be drawn. On the other hand, the files are only one part of the legal proceedings: there are few details about these relationships. It can only be assumed that most creditors distrusted bankrupts. 13 Consequently, the goal of this section is to deduce how the self-interest of the agents was shaped through accounting.
In a few marginal cases, accounting does not appear to exercise any role at all. By disappearing, the bankrupt may have hoped to escape the long arm of the law (see files 3222 and 3223 in 1877). The second expedient consisted of the bankrupt postponing repayment of some of his debts in the hope of receiving, in the meantime, monies enabling him to avoid bankruptcy. This was the case in 1847 with Mrs Dalet (file 7281) who, just before declaring bankruptcy, chose to repay some of her creditors.
Nevertheless, in the majority of cases, proceedings moved forward more routinely. At this stage it is possible to identify two kinds of accounting self-interest.
The first was adopted by a majority of bankrupts at the beginning of the period studied, but subsequently petered out. It consisted in presenting oneself in the best possible light. To this end, the bankrupt would follow the proceedings as conscientiously as possible. He declared voluntary bankruptcy – that is, without waiting for the creditors to demand it – and then filed his accounting balance sheet before the legal deadline (three days), allowing the receiver to conduct his inventory without attempting to conceal any assets. His valuation was often intentionally optimistic and aimed to show creditors that they would recover a sizeable dividend.
The second example of accounting self-interest consisted of declaring bankruptcy as late as possible once assets had been depleted and were virtually worthless. Low asset values would prevent the receiver from being paid his expenses and the latter’s investigations were therefore promptly brought to an end with the magistrate quickly ruling to halt proceedings due to insufficient assets. Such a strategy, whereby the brevity of proceedings would enable the bankrupt to conceal any possible errors or frauds, was clearly to the detriment of creditors.
Before examining the advantages and disadvantages of these strategies in greater depth, it is apposite to return to the balance sheets filed by the bankrupts in the sample. The first factor to consider is a comparison between the balance sheets filed and the outcomes of the proceedings (see Table 6).
Proportion of bankrupts filing balance sheets compared with the outcome of the proceedings (1847–1887).
Note: a This figure indicates that 89 per cent of bankrupts who obtained a settlement filed their balance sheet.
Table 6 gives a clear idea of the influence that the filing of a balance sheet could have on the outcome of the proceedings (Hautcoeur and Levratto, 2006: 25). First and foremost, the filing of a bankruptcy balance sheet attested to the bankrupt’s ability to promptly draw up a balance sheet from his bookkeeping (within three days). The speed with which he did this would seem to demonstrate his integrity and was a factor in his favour in deliberations regarding his future solvency. It is hardly surprising that, as a result, the chances of being granted a settlement were significantly higher when the bankrupt voluntarily filed his own bankruptcy balance sheet.
Strategies adopted by bankrupts seeking to present themselves in the best possible light did not end here. For a large number, this initial balance sheet presented a good opportunity to lead creditors to believe that their dividend would be sizeable. 14 Table 7 shows the asset/liabilities ratios calculated on the basis of initial balance sheets.
Rankings of bankruptcies according to the assets/liabilities ratios declared by bankrupts (1847–1887).
Tables 6 and 7 provide a good illustration of the accounting self-interest of most bankrupts until 1867 – the year in the period of the present study representing the moment when strategies shifted. Between a third and a half of all files presented an asset/liabilities ratio higher than 75 per cent.
From 1867 onwards, the proportion of bankrupts who filed balance sheets fell dramatically, as shown in Table 3. How can this shift be explained? Bankruptcy statistics provide one indication that may explain this change in practices – see Table 8.
Number of bankruptcies in the Paris region.
Note: a The average for the decade is calculated from the four years before and the four years after the stated year. For 1847, we calculated the average of the years 1843, 1844, 1845, 1846, 1848, 1849, 1850 and 1851.
Source: Jobert (1991: 175–176).
These figures show a rise in the number of bankruptcies over the relevant period. It should be noted in passing that 1847 was an exception compared to the rest of the 1840s. In 1846, there were 939 bankruptcies in the Paris region (the average for the decade), and only 634 in 1848. In other words, the figure of 1,330 was significantly above the average, which was just below one thousand. The average for the decade confirmed the trend: the number of bankruptcies rose over this period. Between the decades 1840 and 1850 and 1860, 1870 and 1880, a doubling in the number of bankruptcies is observed, in line with Lemercier’s (2008: 66) observation that over the same period the Courts of Commerce became clogged up.
The quality of the work carried out by receivers probably suffered as a result. They had less time to examine each case and took less interest in cases where they would not be able to cover their costs. As a consequence, not filing a balance sheet became much more common, and a number of bankrupts no longer bothered to do so, safe in the knowledge that proceedings would promptly be halted due to insufficient assets. It is therefore unsurprising to see numbers of both settlements and liquidations decline from 1867 onwards. Failing to file a balance sheet within three days no longer seems to have been viewed as a major offence, as the receiver Heurtey implied in 1867:
Despite the understanding that Frenal, as a bankrupt trader, did not declare a suspension of payments to the Court of Commerce within the deadline set by the law, that he did not conduct a thorough inventory and nor did he keep consistent and complete books, there is no reason in the circumstances to declare him guilty of banqueroute simple. (file 7350)
Even when balance sheets were not filed by bankrupts, in most cases balance sheets were still prepared at the outset of bankruptcy proceedings – balance sheets co-signed by the bankrupt and the receiver.
In practice, valuations in this balance sheet barely differed from those presented in the settlement balance sheet. It is reasonable to assume that the latter was drawn up after the receiver carried out the inventory. This article has, therefore, ignored the former balance sheet, which in no way equates to the balance sheet filed by the bankrupt within three days of declaring bankruptcy.
This section has highlighted two forms of accounting self-interest: the first, prevalent in the earlier period, consists of the bankrupt overvaluing his assets, whilst nonetheless scrupulously respecting the law. The second accounting self-interest goes hand in hand with the increase in the number of bankruptcies. Bankrupts then envisaged using accounting to abridge the proceedings by letting it be understood that their assets were so low that there would not be sufficient to repay the creditors.
The accounting self-interest of the receiver
The accounting self-interests of the receiver are far more difficult to ascertain. However, it can be safely assumed that the receiver had an interest in the smooth running of proceedings so that no party could lodge any complaint. Several facts corroborate this assumption.
Following the Law of 1838, the Code de Commerce did not explicitly set out the organisation of the profession of receiver. It confined itself to banning any of the bankrupt’s relatives or allies from being appointed as the receiver (Article 463) and left it to magistrates at the Court of Commerce to make such appointments after consultation with creditors. With respect to the Paris courts, the 1866 Manual of the Court of Commerce provides an elucidating response to the question of why it was so difficult to find competent receivers when between 1,000 and 2,000 bankruptcies were filed each year:
[Receivers] in Paris today constitute a serious profession and those who are admitted to exercise this profession deserve, for their worthiness and for the services they render, the esteem of both the traders and the magistrates who see them at work on a daily basis … Indeed, who could reasonably claim that in Paris, for instance, the fifteen to eighteen hundred bankruptcies declared each year could be properly managed and administered by people unknown to the courts and who would hold no liability themselves? (Teulet and Camberlin, 1866: 192)
As the profession became more organised, it might be expected that this would lead to the observation of a correspondingly gradual change in practices, with the number of appointed receivers falling, and with each handling more and more cases. Table 9 enables testing of this assumption on the basis of the sample.
Number of bankruptcy cases handled by receivers (1847–1887).
The number of receivers fell significantly: from 39 in 1847 to 22 by 1857, and subsequently remained around this level. Moreover, handling 100 files did not take the same amount of time from one year to the next. Naturally, the corollary of this shift was a rise in the number of cases handled by each receiver. In 1847, no receiver handled more than six cases, for a total duration of 31 days. In 1877, for a comparable amount of time (27 days), one receiver handled 14 cases, another 12, and two others eight each. Since receivers’ time was not scalable, 15 this would mean that the time spent by each receiver on each case tended to fall – see Table 10.
Average time spent on a case by each receiver (1847–1887).
Under such conditions, it is not surprising that there was a significant rise in the number of proceedings halted due to insufficient assets. With the number of cases rising incessantly, it made sense to spend time on bankruptcies that had the best chance of generating sufficient assets.
The hypothesis that the profession became more organised would, therefore, seem to be confirmed. This professionalisation was rather limited and meant that being a receiver became a paid job and not an unpaid position, as it had been before the Law of 1838. However, there was no professional organisation for receivers, such as the creation of an order. In the 1980s, this lack was one explanation for the disappearance of the profession (Dezalay, 1989).
The arguments expounded here appear convincing. Professionalising the receiver’s function led to improvements in their competence. Multiplying the number of cases they handled gave them greater knowledge of the behaviours of both bankrupts and creditors. The trade-off for receivers’ experience was, of course, less time spent on each case, as detailed above.
The second argument in favour of professionalisation was the power this gave magistrates. Since the receiver had no income other than revenues generated by the cases to which he was assigned, he had to comply with the rules in order to acquire a good reputation. This raised the question of how this might be achieved. The first step was to show integrity: no embezzlement.
Indeed, the Manual of the Court of Commerce (pp.203–206) reveals that the Paris Court of Commerce set up a central accounting system at the seat of the court itself. It operated in the following manner: the receiver submitted for each working day an account of the operations carried out and paid any monies collected into the “Caisse des consignations” – a public centralised fund. It was only at the end of the bankruptcy proceedings, after verification by the central accounting office and after receiving a full discharge, that receivers received payment for their expenses. Although this mechanism did not prevent all fraud, it can reasonably be assumed to have limited it.
After verification by the magistrate and by the central accounting office, the receiver was subject to a third verification – by the creditors themselves. Article 537 stipulated that creditors had to grant the receiver a full discharge with respect to his administration. Refusal to grant such a discharge would mean that it was impossible to reimburse any expenses tied to bankruptcy proceedings. It is likely that such a refusal (especially if repeated) would tarnish the receiver’s reputation in the eyes of the magistrates.
It is in the light of this context that the valuation practices in bankruptcy proceedings examined previously should be considered. Application of the principle of prudence and/or the market liquidation values meant that receivers presented very low asset values in settlement balance sheets. It was consequently easier for receivers to sell off these assets at a similar (if not higher) value. As a result, the final dividend would be at least equal to, if not higher than, the expected dividend. The receiver could then present himself to creditors as a good administrator and obtain a full discharge more easily.
Low asset valuations had another advantage. By announcing an insignificant dividend, the receiver deterred some creditors from making their claims. It was not uncommon to see liabilities fall between settlement and liquidation, enabling the dividend to rise automatically (file 1516).
This section has highlighted the fact that the behaviour of the receiver cannot be explained only by the move to professionalisation. Like the bankrupt, he was part of judicial proceedings and needed to be very cautious with respect to the creditors and the judge.
The construction of the self-interest of creditors
For the bankrupt and the receiver it was possible to link accounting to the idea of self-interest since each agent drew up a balance sheet, at the minimum. For the creditor, it is not a case of revealing accounting self-interest, but of seeing how the accounting self-interest of the other agents (bankrupt and receiver) influenced the representations, the self-interest and, in the long run, the behaviour of the creditors.
Unlike receivers, creditors are a varied group and their differences are many (profession, status, and importance of their debts, etc.). Nevertheless, a study of the bankruptcy files allowed identification of their reactions to the self-interest of other agents. This section analyses more precisely how the creditors’ choice between settlement and liquidation might be influenced by the accounting self-interest of bankrupt and receiver.
In practice, the receivers’ role only emerged at the time of settlement, at which point they only had two options: they could either grant a settlement, accompanied most often by a discount in their credits and a rescheduling of the bankrupt’s debt repayments, or reject a settlement, in which case proceedings would move towards liquidation. The main question this raises is the extent to which the receiver’s report may or may not have influenced the creditors’ decision and, secondly, whether or not creditors always followed the more or less explicitly formulated advice given by receivers.
The examination of settlement balance sheets revealed that the receiver’s position was never asserted. It is therefore difficult to weigh up his opinion in a quantifiable way. The data presented in Table 11 was collected from the sample.
The influence of receivers over creditors’ choice of the proceedings to follow (1847–1887).
Note: a The significant rise in the number of proceedings halted due to insufficient assets explains the drop in the number of files available for use.
To better understand how such recommendations worked, take the example of the year 1847. To elicit a settlement from creditors, receivers used two arguments: first of all, the good faith of bankrupts (or at least the fact that they could not be accused of any breach of faith) and, secondly, that opting for liquidation would entail legal proceedings resulting in a very small dividend. Such arguments appear to have been convincing since, in most cases, creditors chose settlement. Conversely (and less commonly), receivers could also make remarks designed to temper creditors’ decisions. The receiver could remind creditors that his examination of the accounts did not allow verification of the bankrupt’s statements (files 7234, 7271, 7285 and 7286), implying possible dishonesty even when it could not be proven. In some cases, criticism was even more explicit (files 7287 and 7291).
In total, in 58 cases (36 + 22) out of 98 (see Table 10), the creditors tended to follow the receiver’s advice. Yet such a scenario was far from being the rule. Once again, it is necessary to delve further into the proceedings to understand the exceptions to this rule. The bankrupt had to make a request to his creditors for a settlement. If he did not do so, then proceedings would automatically move to liquidation (file 1545).
Nevertheless, in the majority of cases, bankrupts did submit proposals to their creditors. They had to specify several points: the percentage discount they were requesting, the waiver (or not) of some or all assets and, finally, a payment rescheduling agreement. To be legally binding, the settlement “can only be reached with the agreement of a number of creditors forming the majority and, furthermore, representing three-quarters of the total credits as verified, stated or provisioned for” (article 507).
It may be assumed that many creditors preferred liquidation because it seemed the best way to recover a part of their credits, whereas settlement constituted a more risky option. From this viewpoint, choosing a settlement may be viewed as pertaining to an investment decision. Moreover, the double-majority system stipulated in Article 507 made choosing a settlement more difficult than rejecting it.
Above and beyond these reasons, a hierarchy among creditors could also weigh on the outcome of the proceedings. Some creditors received preferential treatment in the eyes of the law (i.e. those holding mortgages or pledges). Other, so-called ordinary creditors were unlikely to recover anything because low asset values often barely covered payments to preferred creditors (files 7327 and 7375).
The nature of assets also influenced creditors’ decisions. The more liquid the assets (i.e. they could be sold off at a minimal loss), the more creditors preferred liquidation, ensuring short-term recovery of a significant part of their credits. Conversely, when assets were almost entirely made up of the fonds de commerce (as was the case in the bankruptcy of the pastry baker Lafouge (file 13691)), for instance, creditors had an interest in granting a settlement. The difference between what a skilled trader could generate from a going concern and its liquidation price appears to have been fairly significant.
In the case of creditors, therefore, self-interest emerges. The role of accounting is no longer to merely make manifest the self-interest of an agent, but to influence the behaviours of another agent.
Concluding remarks
This article highlights several key points for researchers in accounting and business history.
The first point concerns the practices of accounting valuation in nineteenth-century France. In particular, historic cost without depreciation was in widespread use by the smallest companies. Therefore, this research reveals differences between companies and small and medium traders in their accounting valuation practices.
The second point relates to the complexity of the notion of value. There are several possible values for any given asset. None is, in and of itself, truer (or more false) than the others; it depends primarily on the purpose assigned to it. To use contemporary terminology, historic cost, market value or discounting future revenues emerge as part of a range of suitable options. Added to these issues of valuation are persistent doubts as to the good or bad faith shown by various actors, each suspected of adopting the valuation that best serves his interests.
The third point concerns the origins of financial reporting. Bankrupts tried to use (sometimes clumsily) their balance sheets as a tool for financial reporting to receivers and, above all, their creditors. The goal was, for instance, to influence the decision of creditors and to obtain a settlement rather than a liquidation. A similar analysis can be made for the receiver: he had some particular interests and his accounting valuations reflected this. His position resembles that of the auditor (who was emerging in nineteenth-century France). Receivers’ behaviors seem to be closely linked with their nomination proceedings and their payment. These types of practices must be set alongside the development of limited companies in France (Lemarchand, 1993a) and in Anglo-Saxon countries (Ó hÓgartaigh, 2009). Both for limited companies and bankruptcies, questions concerning the valuation of assets are also central. In each case, the process of valuation may be understood as the defence of the special interests of agents, bringing us back to the idea of self-interest.
The fourth point is about the construction of the accounting self-interest of agents and it has been shown how accounting, as a calculative tool, can become a means of expressing self-interest. The process of bankruptcy, therefore, reveals a double phenomenon: on the one side its evolution reveals the professionalisation highlighted by other commentators (Coquery and Praquin, 2008) and, on the other, individual behaviours are seen that use accounting tools to shape the representations of the other agents. This is not really surprising for the bankrupt, but it is for the receiver who was supposed to be neutral in the bankruptcy proceedings. This article has brought to the fore several ways for each agent to express his own self-interest through accounting valuation, and it is obviously impossible to sum up all these cases here.
Until now, bankruptcy archives were mainly used as a way to understand economic, legal or social history. Nevertheless for accounting historians too, bankruptcy archives are a very rich source. Moreover, a good understanding of accounting practices is very useful in understanding what was at stake between the bankrupt, the receiver, and the creditors.
Footnotes
Acknowledgements
The author would like to thank Garry D Carnegie, John Richard Edwards, Richard Fleischman, Yannick Lemarchand, Lee D Parker, Stephen P Walker, and the two anonymous referees for their useful comments and suggestions.
Declaration of conflicting interest
The author declares that there is no conflict of interest.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.