Is data being giv­en enough con­sid­er­a­tion in account­ing procedures?

Europe is evolv­ing con­tin­u­ous­ly from a man­u­fac­tur­ing to a ser­vice-ori­ent­ed soci­ety. The dig­i­tal rev­o­lu­tion is also pro­gress­ing. Are cur­rent account­ing and tax laws armed and ready for these inter­re­lat­ed challenges?

In her Duet Inter­view with legal and data expert Dr Angel­i­ca M. Schwarz, Dr Cal­daro­la, author of Big Data and Law, dis­cuss­es account­ing and tax issues in con­nec­tion with data and data-dri­ven busi­ness models.

The view con­cern­ing the mar­ket cap­i­tal­iza­tion of impor­tant stock index­es illus­trates our devel­op­ment from a clas­sic indus­try to a soci­ety focussing on ser­vice, high tech and knowl­edge, a change which high­lights data’s role as an intan­gi­ble asset and strate­gic fac­tor for suc­cess. The mar­ket val­ue of equi­ty cap­i­tal in those index­es, mir­ror­ing the expec­ta­tion of investors, exceeds the statu­to­ry declared book val­ue of the equi­ty cap­i­tal by a num­ber of mag­ni­tudes. Rea­sons for this dis­crep­an­cy between mar­ket and book val­ue include the use of undoc­u­ment­ed, intan­gi­ble finan­cial assets as well as the out­look for the future by ana­lysts. There is the assump­tion that the mar­ket cap­i­tal­iza­tion of the US stock index Stan­dard & Poor 500 demon­strates a clear trend towards an increas­ing share of intan­gi­ble assets. When observ­ing the mar­ket-to-book ratio of com­pa­nies with infor­ma­tion-based busi­ness mod­els, one can expect that the dis­crep­an­cy will keep on increas­ing. What is your opinion?

Dr Angel­i­ca M. Schwarz: In my view, we should be ask­ing the ques­tion why cer­tain intan­gi­ble assets are not reflect­ed in the bal­ance sheets at their mar­ket val­ues. The answer is sim­ple: Because com­mon account­ing reg­u­la­tions do not pro­vide for their inclu­sion. Let us con­sid­er the fol­low­ing example:

The Inter­na­tion­al Finan­cial Report­ing Stan­dards (IFRS) are based on the true and fair view prin­ci­ple. What does this mean con­crete­ly? The prin­ci­ple states that the annu­al finan­cial state­ment should enable the read­er or stake­hold­er to gain an insight into the actu­al posi­tion of assets and earn­ings of the com­pa­ny in ques­tions. The finan­cial state­ment should there­fore mir­ror the trans­ac­tions in a trust­wor­thy man­ner, name­ly, to con­vey the actu­al finan­cial sit­u­a­tion of a com­pa­ny. In con­trast to Swiss account­ing reg­u­la­tions, the IFRS fol­lows a nar­row inter­pre­ta­tion of the pru­dence prin­ci­ple. In gen­er­al, pru­dence is defined as the exer­cise of cau­tion when mak­ing judge­ments under con­di­tions of uncer­tain­ty. Thus, under the pru­dence con­cept, one should fol­low a con­ser­v­a­tive approach in record­ing the amount of assets, and not under­es­ti­mate lia­bil­i­ties. Even though the pru­dence prin­ci­ple is defined dif­fer­ent­ly, both the Swiss account­ing rules as well as the IFRS require fur­ther pre­req­ui­sites when it comes to self-cre­at­ed intan­gi­bles. The rea­son for such reluc­tance is that intan­gi­ble assets gen­er­at­ed by the com­pa­nies them­selves have not passed a mar­ket test and are there­fore deemed to be uncer­tain. Even though a val­u­a­tion may indeed be pos­si­ble (e.g., based on com­par­a­tive fig­ures from sim­i­lar trans­ac­tions), such assets have not been acquired on the mar­ket and, there­fore, no direct fair mar­ket val­ue (in the sense of a price) is avail­able. Thus, indi­vid­u­al­ly gen­er­at­ed intan­gi­ble assets have to at least pass the research and to have reached the devel­op­ment phase before a cap­i­tal­iza­tion is pos­si­ble from an account­ing perspective.

Once indi­vid­u­al­ly gen­er­at­ed intan­gi­ble assets have over­come this hur­dle and can be reflect­ed in the bal­ance sheet, the ques­tion con­cern­ing under which val­ue they should be list­ed remains open. The ini­tial assess­ment of such resources is one of cost (e.g., pro­duc­tion costs). Includ­ed are all attrib­ut­able direct costs nec­es­sary for gen­er­at­ing said asset. Here­in lies the prob­lem, in my opin­ion, because for the respec­tive com­pa­ny the pro­duc­tion costs are often low­er than the actu­al intrin­sic val­ue of the intan­gi­ble asset(s). This holds espe­cial­ly true if the Big Data archi­tec­ture (once imple­ment­ed) can be used over many years to reduce the com­pa­ny’s costs. I use the term “actu­al” delib­er­ate­ly. As men­tioned in the begin­ning, does not the true and fair view prin­ci­ple require that the actu­al posi­tion of assets and earn­ing of the com­pa­ny be reflect­ed in the bal­ance sheet? In my view, there is a cer­tain dis­crep­an­cy reflect­ed in how the prin­ci­ple is adhered to.

Com­ing back to the ques­tion you posed at the begin­ning: Yes, there is a dis­crep­an­cy in rela­tion to the mar­ket-to-book ratio and the dis­crep­an­cy will con­tin­ue to increase if com­pa­nies decide increas­ing­ly on dig­i­tal busi­ness mod­els. From an account­ing per­spec­tive, dig­i­tal assets gen­er­at­ed by com­pa­nies them­selves (such as dig­i­tal data of third par­ties (i.e. data sub­jects) col­lect­ed and processed by the com­pa­ny itself) rep­re­sent intan­gi­ble assets and as long as they can “only” be priced at pro­duc­tion cost, we will con­tin­ue to observe a dis­crep­an­cy between mar­ket and book val­ue. If this is in line with the true and fair approach and if the annu­al finan­cial state­ment can still serve as a reli­able basis for cap­i­tal mar­ket par­tic­i­pants when mak­ing busi­ness deci­sions is anoth­er question.

Does this dis­crep­an­cy also apply to the account­ing of data?

This is espe­cial­ly true when talk­ing about whether or not data can be list­ed as a busi­ness asset in the bal­ance sheet. In my doc­tor­al the­sis I present the view that data – whether indi­vid­u­al­ly gen­er­at­ed or pur­chased – may ful­fil the con­di­tions as pro­vid­ed by the IFRS in order to list them on the bal­ance sheet. The imple­men­ta­tion of a Big Data project can be very expen­sive. On the one hand, the com­pa­ny requires the nec­es­sary IT-infra­struc­ture (unless a cloud infra­struc­ture is used) and, on the oth­er hand, the cru­cial know-how. Raw data needs to be trans­formed into a cer­tain for­mat so that it can be analysed. This “refine­ment“ process takes on a cer­tain sig­nif­i­cance if data come in dif­fer­ent for­mats from var­i­ous sources and enti­ties of a multi­na­tion­al and need to be reduced to a com­mon denom­i­na­tor. The cost for such a Big Data archi­tec­ture is, from what I have seen, like­ly to be far below the actu­al val­ue of the data. The ini­tial recog­ni­tion accord­ing to IFRS for self-gen­er­at­ed data is there­fore cost-ori­ent­ed and not ben­e­fit-ori­ent­ed. The recog­ni­tion of fair mar­ket val­ue will in most cas­es not apply because an active and iden­ti­fi­able mar­ket is miss­ing. Trade with data exists but it is ques­tion­able whether such trade can already be iden­ti­fied to a suf­fi­cient degree. This, of course, may change in the future and I would not exclude that at some point the trade of data sim­i­lar to mar­ketable securities.

And what about acquired data?

The ini­tial recog­ni­tion of acquired data is based on the acqui­si­tion costs. The sit­u­a­tion here is dif­fer­ent because the data in ques­tion have, in this par­tic­u­lar con­stel­la­tion, over­come a mar­ket test and, among inde­pen­dent third par­ties, the pur­chase price cor­re­sponds usu­al­ly to the fair mar­ket value.

In your opin­ion, data can ful­fil the require­ments for being list­ed on the bal­ance sheet from an IFRS per­spec­tive. Some argue, how­ev­er, that data can­not be at the dis­pos­al of a com­pa­ny and, thus, can hard­ly be sub­stan­ti­at­ed. What is your view?

One should under­stand that the IFRS (so too the Swiss reg­u­la­tion) fol­low an eco­nom­ic approach when deal­ing with the ques­tion whether or not a cer­tain asset is at the dis­pos­al of the com­pa­ny. The require­ment that data must be at the dis­pos­al of the com­pa­ny does not nec­es­sar­i­ly pos­tu­late legal own­er­ship. Instead, this pre­req­ui­site is to be under­stood as an eco­nom­ic pow­er over the data, which can be deter­mined based on cri­te­ria such as risk, oppor­tu­ni­ty, ben­e­fit, lia­bil­i­ty, etc. Fur­ther­more, in order for data to be con­sid­ered as being avail­able for exploita­tion by a com­pa­ny, the com­pa­ny needs to have phys­i­cal or vir­tu­al access to the data pool. How­ev­er, it can­not be pre­sumed that the data car­ri­er itself must also be con­trolled at some lev­el by the com­pa­ny in ques­tion; oth­er­wise, data that is stored in an exter­nal­ly rent­ed cloud could nev­er be allo­cat­ed to a data-dri­ven cor­po­ra­tion using this kind of a tool.

Since eco­nom­ic pow­er and not nec­es­sar­i­ly legal pow­er is the rel­e­vant aspect, the ques­tion whether legal own­er­ship of data is even pos­si­ble is of a sec­ondary nature. Legal own­er­ship can, how­ev­er, cer­tain­ly be used as an indi­ca­tion of control.

Data also car­ries risks. The legal pro­cess­ing of data has not been ful­ly clar­i­fied. How should com­pa­nies han­dle such risks when pro­cess­ing data and, at the same time, infring­ing upon data pro­tec­tion law? I am think­ing of era­sure claims in accor­dance with data pro­tec­tion laws.

Data is report­ed on the asset side on a bal­ance sheet where­as risk is list­ed on the lia­bil­i­ty side. In my view, it is impor­tant to under­stand that a com­pa­ny col­lect­ing and analysing data can still ben­e­fit from such activ­i­ties even though they may infringe upon data pro­tec­tion law reg­u­la­tions. That, of course, does not mean that I give a com­pa­ny carte blanche to do what­ev­er they desire with data. What I would like to point out is that the ques­tion whether or not data can be list­ed on the bal­ance sheet needs to be eval­u­at­ed inde­pen­dent­ly of whether and to what extent the com­pa­ny should make pro­vi­sions in order to cov­er pos­si­ble future lia­bil­i­ties. The claim for era­sure, how­ev­er, may be con­sid­ered when eval­u­at­ing whether and to what extent a val­ue adjust­ment seems necessary.

Multi­na­tion­als often imple­ment their big data strate­gies for the whole group. How do you assess data trans­fer among com­pa­ny subsidiaries?

With regard to affil­i­at­ed com­pa­nies, the ques­tion aris­es whether and which price should be paid.

The ques­tion of “whether” can be answered more eas­i­ly. Tax law requires that a trans­ac­tion between affil­i­at­ed com­pa­nies needs to be real­ized at arm’s length. This means that the com­pa­ny must pay the price that a third par­ty would have paid. The log­ic behind this is to pre­vent prof­it shift­ing. Oth­er­wise, a sub­sidiary in a low-tax coun­try could be tempt­ed to make a high prof­it by sell­ing its data at a price beyond mar­ket val­ue. If the con­tract­ing par­ty is a sub­sidiary in a high-tax coun­try, it could have an inter­est in pay­ing such a high price, as the com­pa­ny can expense the costs via the prof­it and loss state­ment which, in turn, reduces its tax base. The real­i­ty today is that data trans­fers among group com­pa­nies exist while, how­ev­er, usu­al­ly no price at all is being paid. If we treat data as an asset, why should a com­pa­ny not com­pen­sate the data it receives from oth­er group com­pa­nies like any oth­er asset or ser­vice as well? That might sound neg­a­tive at first glance but there are also oppor­tu­ni­ties – name­ly where tax plan­ning is con­cerned. Sur­pris­ing­ly, this top­ic has not been of great sig­nif­i­cance either for com­pa­nies or for tax author­i­ties. The ques­tion remains for how long.

If the “whether” has been set­tled, then the next ques­tion is how the price should be deter­mined. Here a mar­ket-val­ue-ori­en­ta­tion is allowed – if not required. Dif­fer­ent to the account­ing reg­u­la­tions, an actu­al val­ue needs to be iden­ti­fied. This is the point where the inter­play between tax and account­ing reg­u­la­tions becomes very inter­est­ing: A price assessed for tax rea­sons may have a direct influ­ence on the account­ing of a com­pa­ny. My view here is that any prices which are in accor­dance with the arm’s length prin­ci­ple – although they are paid among affil­i­at­ed com­pa­nies – should be accept­ed as the pur­chase price in an indi­vid­ual finan­cial statement.

My favorite quote is:

“The art of col­lect­ing tax­es is to pluck the goose with­out it screaming”.

Max­im­i­lien de Béthune, Duc de Sully

Dr Schwarz, thank you for shar­ing your insights on data and account­ing reg­u­la­tions, trans­fer pric­ing and tax.

Thank you, Dr Cal­daro­la, and I look for­ward to read­ing your upcom­ing inter­views with rec­og­nized experts, delv­ing even deep­er into this fas­ci­nat­ing topic.