• The tax court may not simply replace a purchase price allocation agreed in the purchase agreement (land vs. building) with the allocation determined using the BMF working aid, even if the contractual allocation falls significantly short of the actual values.
• The BMF working aid does not guarantee an apportionment according to real market values because it is methodically narrowed down to a simplified asset value method and does not take into account a local or regionalization factor for the building value; as a result, building values tend to be set too low and land shares too high in high-price regions.
• In the event of a disputed property valuation, the tax court must generally obtain an expert opinion from a publicly appointed and sworn expert, unless it has its own proven expertise in exceptional cases.
• In this specific case, the BFH overturned the ruling of the Berlin-Brandenburg tax court and referred the case back for a new hearing: although the contractual allocation was not viable, the replacement by the BMF working aid was also inadmissible.

Facts of the case

I.

1. In dispute is the allocation of the purchase price for a rented condominium to the building and the land for the purpose of determining the depreciation for wear and tear (AfA).

2. The plaintiff and appellant (plaintiff), a real estate community, acquired a 38.83 sqm condominium in A-Straße in Z by notarized contract dated 26.04.2017, combined with a co-ownership share of 38.577/1,000 in the property with a size of 1,185 sqm and the special right of use of a cellar room and a double garage space. It is a rented one-room apartment on the second floor of an apartment building completed in 1973. According to the purchase agreement, the purchase price was € 110,000, of which € 2,642.05 was attributable to the share of the maintenance reserve as at 30.04.2017. The purchase agreement also contained the following provision: "The purchase price includes the consideration for the pro rata value of the property, which the parties estimate at € 20,000.00 to the best of their knowledge at the time the agreement was concluded." The acquisition costs, including ancillary costs, amounted to € 118,002.

3. In its declaration of assessment for the 2017 year of dispute, the plaintiff took into account a depreciation assessment basis of € 96,547.47. It based this on the purchase price allocation of the notarized purchase agreement and calculated a building share of 81.81 % (20,000/110,000).

4. The defendant and appellant (the tax office --FA--), on the other hand, calculated a building share of 27.03 % on the basis of the "Arbeitshilfe zur Aufteilung eines Gesamtkaufpreises für ein bebautes Grundstück (Kaufpreisaufteilung)" provided by the Federal Ministry of Finance (BMF) on the Internet (www.bundesfinanzministerium.de). This was based on a land value of € 77,713 (1,185 sqm [area] x € 1,700/sqm [standard land value] x 38,577/1,000,000 [co-ownership share]) and a building value of € 28,782 (39 sqm [living space] x € 738/sqm [standardized production costs --THK--]). Accordingly, the acquisition costs of € 118,002 were divided up and subjected to depreciation in the amount of € 31,896 in the 2017 assessment notice on the separate and uniform determination of tax bases dated 05.10.2018.

5. In the appeal proceedings, the claimant argued that the land value stated in the purchase agreement deviated from the standard land value and that she had comprehensively examined the building, which was constructed by a "star architect", before purchasing the apartment. Not only the condition of the building, but also the layout of the apartment and the room layout, the design of the hallways and the large balconies were decisive for the purchase. On the other hand, the location of the property (between the S-Bahn and the highway, near a square with a high crime rate, cobblestones in front of the door) was not good. The standard land value was therefore unfounded. The share shown in the purchase agreement had been allocated by the parties to the purchase agreement - they were third parties - in a well-considered and appropriate manner. The expenses for the renovation of the bathroom and kitchen alone would be higher than the building value share to be taken as a basis in accordance with the working aid.

6. The FA followed this in the objection decision of 8 May 2019 only to the extent that it now applied a building value share of 30.9 % and thus a depreciation assessment basis of € 36,463 because it had disregarded the underground parking space (€ 5,965, total building value thus € 34,747) when calculating the building share.

7. The plaintiff appealed against this and also claimed that the building value share stated in the purchase contract was still far below the current building production costs of at least € 2,000 per square meter, so that an amount of € 80,000 should be regarded as the minimum value for the building share. According to publications by the Zentralverband des Deutschen Baugewerbes (Central Association of the German Construction Industry), construction costs of € 2,400 per sqm and pro rata land costs of € 600 per sqm are to be assumed for inner-city locations, resulting in a land value share of 20%, as stated in the purchase agreement. It is true that the purchase price allocation in the notarized purchase agreement was initiated by the buyer and that the seller had no reason to oppose it. However, the allocation was based on the fact that a calculated return of 6.71% - based solely on the value of the building - was to be achieved.

8. In the ruling published in Entscheidungen der Finanzgerichte (EFG) 2020, 182, the tax court dismissed the action as unfounded. It essentially stated that the contractual purchase price allocation did not reflect the actual value ratios in the case in dispute. It considered the working aid for the valuation, in particular for the determination of the real value of the building, to be suitable in principle, attributed great indicative significance to its results in order to be able to refute the market appropriateness of the contractually agreed purchase price allocation in the event of a significant deviation, and considered it to be a suitable estimation aid in the event of refutation.

9. The plaintiff's appeal is directed against this. In addition, it asserts that the working aid does not accurately represent the actual value ratios by means of standardized estimates in the area of building value. Even the legislator - as can be seen from § 7b of the Income Tax Act (EStG) - does not consider construction costs of € 2,000 per square meter for simple living spaces to be feasible; the subsidy limit was therefore set at € 3,000 per square meter - with an imminent adjustment to € 3,500 per square meter. However, the construction costs taken into account in the present case in accordance with the Sachwertrichtlinie (SW-RL) of 05.09.2012, published in the Bundesanzeiger, Amtlicher Teil, 18.10.2012 B1, amounted to € 1,405 per sqm (without age value reduction). Furthermore, the tax court had disregarded the "value of living". The determination of the building share by the tax authorities was based unilaterally on the asset value. The plaintiff's rental income was not offset by any income-related expenses corresponding to the residential value.

10. The plaintiff applied for the judgment of the Berlin-Brandenburg tax court dated 14.08.2019 to be set aside and for the assessment notice dated 05.10.2018 in the form of the objection decision dated 08.05.2019 to be amended to the effect that the income from the letting of the A-Straße property was determined on the basis of an AfA assessment basis taking into account a building share of 81.8 %.

11. The FA requested that the appeal be rejected.

12. It essentially argues that the lower court correctly applied the case law on the allocation of the total purchase price to the acquired assets. There were considerable doubts about the agreed purchase price allocation, which the plaintiff had not been able to refute with other qualified evidence. Against this background, the review of the purchase price allocation was justified. The working aid issued by the BMF was a suitable instrument for this purpose. In particular, it goes beyond a mere determination of the standard land values and their comparison with the purchase contract land share. It is based on the regulations for determining the market value on the basis of the Federal Building Code (BauGB), in this case the real value method in accordance with the Real Estate Value Ordinance (ImmoWertV) of 19.05.2010 (BGBl I 2010, 639). Contrary to the plaintiff's statements, the building value is determined on the basis of the normal production costs (NHK) 2010, which are used as the starting value. Finally, the price is indexed in accordance with the construction price index (in the case in dispute at 116.8% of the NHK 2010). This results in an adjustment to current values. In this way, the current standard land values are also compared with adjusted construction costs.

13. The plaintiff wrongly assumes that the price increase of residential properties in a certain location resulting from an increased demand and thus a shortage would practically only lead to an increase in the land value and thus to a corresponding shift in the purchase price allocation of existing properties due to the calculation system of the working aid. Only a market adjustment of the (provisional) material values was dispensed with, as this affects the land on the one hand and the building on the other in the same proportion. The sum of the individual values determined (provisional material values) therefore did not show the market value without a market adjustment. However, the apportionment ratio does not change.

14. Contrary to the plaintiff's view, the THK taken into account in the purchase price allocation was appropriate. According to the indexing of the NHK from 2010 of € 1,405 per sqm to the year 2017, production costs of € 1,642 per sqm of living space would result for a building with the standard of the year 1973. This does not contradict the maximum subsidy limit of € 2,000 per sqm and the upper construction cost limit according to Section 7b EStG of € 3,000 per sqm for a current construction standard, but is plausible.

15. The working aid thus makes it possible to either carry out the purchase price allocation oneself in a standardized procedure or to check the plausibility of a contractual purchase price allocation. It is a qualified estimate, which is inherent in the fact that not all individual circumstances can be assessed. However, the taxpayer can assert special features that justify a different valuation, e.g. by means of an expert opinion. This is a corrective to balance out price differences and injustices. In this way, a proportionate and predictable handling of the valuation for the parties involved was created on the one hand and a possibility to take special individual circumstances into account on the other.

16. The BMF has joined the proceedings and supports the FA. The BMF essentially points out that the review of a contractual purchase price allocation, as well as the allocation itself, can be carried out by an expert opinion. In mass tax proceedings, however, not all purchase price allocations could be reviewed by building experts; rather, in view of the high number of real estate transactions, a weighted approach was required (Section 88 (2) of the German Fiscal Code --AO--), so that the building experts would be consulted in the case of special buildings and in cases of doubt. In all other cases, the assessment offices would have to carry out or check the purchase price allocation - with reference to the working aid. The working aid ensures - as demanded by the state courts of audit (Annual Report of the State Court of Audit of Lower Saxony for 2017, p. 88) - effective and uniform enforcement of the law nationwide. It was downloaded just over 200,000 times from the BMF website in 2019. This -- coupled with the low number of citizen inquiries (approx. 20 in 2019) -- shows the great comprehensibility and acceptance of the tool on the part of taxpayers. It is a qualified estimate that can be refuted by experts and implements the requirements of the ImmoWertV and the SW-RL. On the other hand, the working aid does not claim to replace an expert opinion.

17. The reference to current construction prices is ensured by the annual updating of the working aid, taking into account the price indices for the construction industry published by the Federal Statistical Office. In addition, the reduction in age value can be understood as a market adjustment, unless it concerns a new building.

18. The criticism leveled against the working aid, that in regions with a strong excess demand only the land value and thus the land value share increases and regional differences in construction costs are not taken into account, is misguided. Higher purchase prices did not lead to an increase in the land value share alone due to the mechanism of the working aid (ratio calculation). In addition, the local differences in property price increases can be explained in particular by the location. The increases in construction costs, which deviate only slightly from the general market, are hardly significant compared to the increases in land value. Rather, an increase in the price of land is regularly accompanied by an increase in construction prices, as explained by the tax court. This is reflected in the construction price index. The SW-RL also does not provide for a correction of the NHK 2010 by means of regional factors; there are no (official) nationwide regional construction cost factors. The fact that the results of the working aid can be appropriate for older properties in high-priced large cities, even with a land value share of 80%, can be understood on the basis of the property market report for Munich for 2018 (p. 58 f.). In the case of the dispute, the regionally varying construction price development in the federal state in question is also sufficiently taken into account by the indexing. The deviations between the indices for construction prices of the Federal Statistical Office and Statistical Office X are negligible.

19. The BMF did not file an application.

Reasons for the decision

II.

1. The appeal is well-founded. Pursuant to § 126 para. 3 sentence 1 no. 2 of the Fiscal Court Code (FGO), it leads to the annulment of the contested judgment and to the referral of the case back to the Fiscal Court for a different hearing and decision.

2. The acquisition costs for the building to be determined by the tax court form the basis for determining the depreciation (see 1. below). The lower court correctly did not follow the contractual allocation of the purchase price for the condominium acquired by the plaintiff to land and building (see 2.). However, the tax court was wrong to base its depreciation assessment on the purchase price allocation determined on the basis of the BMF's working aid (see 3.). The matter is not ready for decision (see 4.).

3. 1. the amount of the building depreciation is based on the acquisition costs for the building (§ 7 para. 4 EStG). Their amount forms the basis for the assessment of the depreciation. Its determination is the responsibility of the tax court as the court of fact. This also applies to the allocation of the acquisition costs of the building on the one hand and the associated land on the other (judgment of the Federal Fiscal Court --BFH-- of 16.09.2015 - IX R 12/14, BFHE 251, 214, BStBl II 2016, 397, margin no. 18).

4. a) If a purchase price allocation was made in the purchase agreement, these agreed and paid acquisition costs are generally also to be used as the basis for taxation (established case law, cf. BFH rulings in BFHE 251, 214, BStBl II 2016, 397; from 01.04.2009 - IX R 35/08, BFHE 224, 533, BStBl II 2009, 663; from 18.01.2006 - IX R 34/05, BFH/NV 2006, 1634). Even if the buyer is typically interested in a higher acquisition value of the building with regard to his AfA entitlement and the corresponding apportionment agreement - in favor of the seller - may have an influence on other contractual arrangements that are positive for him, this does not justify a different distribution.

5. However, agreements between the contracting parties on individual prices for individual assets are not binding if there are indications that the purchase price was only determined as a sham (see BFH ruling of 28.10.1998 - X R 96/96, BFHE 187, 450, BStBl II 1999, 217, under B.IV.2.c, para. 33) or the conditions for an abuse of structuring within the meaning of Section 42 AO are met (BFH rulings in BFHE 224, 533, BStBl II 2009, 663; in BFH/NV 2006, 1634, with further references; BFH decision of 04.12.2008 - IX B 149/08, BFH/NV 2009, 365).

6. However, even with an agreement on which taxation is to be based according to general principles, the parties cannot determine the amount of the buyer's tax - specifically its depreciation - in view of the factual nature of the taxation. For this reason, the tax court must examine whether there are any significant doubts about the contractual apportionment when determining the depreciation assessment basis in individual cases (see BFH ruling in BFH/NV 2009, 365, with further references) (BFH ruling of 10.10.2000 - IX R 86/97, BFHE 193, 326, BStBl II 2001, 183; BFH decisions of 24.01.2007 - IX B 84/06, BFH/NV 2007, 1104; of 16.09.2002 - IX B 35/02, BFH/NV 2003, 40). It may not limit itself to reproducing the contractual allocation under tax law, but must verify the result by means of further circumstances, in particular the prices or market values that can be objectively achieved on the market (see BFH decision in BFH/NV 2009, 365).

7. A significant discrepancy with the standard land values does not automatically justify replacing these with the agreed values or estimating the acquisition costs attributable to the land and buildings. It is merely an indication that the contractual apportionment may not adequately reflect the values. Such an indication can be refuted by other indications. The tax court must clarify the overall circumstances of the object of purchase and assess whether particular aspects make the deviation appear plausible. This includes, for example, special features of the building, its original construction costs and any renovations, any limited usability due to existing rental agreements or the residential value of the building in the context of the neighborhood (e.g. street noise, social facilities or special tranquility due to a neighboring green area). At the same time, the tax court must take into account the special criteria of the property, such as a well-kept garden or disturbing trees (BFH ruling in BFHE 251, 214, BStBl II 2016, 397, margin no. 22).

8. A correction of the allocation of the purchase price to land and buildings made by the parties is only necessary if it fundamentally misrepresents the real value ratios and appears economically untenable (BFH ruling in BFHE 251, 214, BStBl II 2016, 397, para. 23, with further references).

9. The tax court has a certain amount of leeway in its overall assessment. In this respect, the general principles of the fiscal court's determination and assessment of the facts apply. In this context, it is not possible to be bound by any estimates made by the FA (BFH ruling in BFHE 251, 214, BStBl II 2016, 397, para. 24).

10. b) If, according to these principles, an agreed purchase price allocation cannot be used as a basis for taxation, the tax court must replace it with an allocation according to the actual market values of the land and building in accordance with its overall assessment of the circumstances. In doing so, the tax court must answer the question of which valuation method is to be used for the purchase price allocation, taking into account the circumstances of the individual case (see BFH decision of 22.10.2007 - IV B 111/06, BFH/NV 2008, 360, with further references; BFH ruling in BFHE 251, 214, BStBl II 2016, 397).

11. c) When allocating a total purchase price, the land and building value must first be determined separately and then the acquisition costs must be divided into acquisition costs for the land and building share according to the ratio of the two value shares. The ImmoWertV can be used to estimate the value of the land and building portion, as it contains recognized principles for estimating the market value of real estate (Senate ruling of 15.11.2016 - IX B 98/16, BFH/NV 2017, 292, margin no. 4; Senate ruling of 15.01.1985 - IX R 81/83, BFHE 143, 61, BStBl II 1985, 252, under 1.b, para. 17, with further references regarding the predecessor regulation in the form of the Value Determination Ordinance of 15.08.1972, BGBl I 1972, 1416, most recently of 06.12.1988, BGBl I 1988, 2209).

12. Accordingly, the market value is to be determined using the comparative value method (including the method for determining the land value), the capitalized earnings value method, the asset value method or several of these methods (section 8 (1) sentence 1 ImmoWertV). The procedures are to be selected according to the type of the object to be valued, taking into account the usual business practices and the other circumstances of the individual case, in particular the available data; the choice is to be justified (section 8 (1) sentence 2 ImmoWertV). Which of these valuation methods is to be used is to be decided according to the actual circumstances of each individual case. The valuation methods are equivalent to each other (see BFH rulings from 27.11.2017 - IX B 144/16, BFH/NV 2018, 218, margin no. 5, and in BFH/NV 2017, 292, margin no. 4, with further references).

13. The determination of the market values is part of the FG's assessment of the facts, which is generally binding for the appellate court (Section 118 (2) FGO). However, the BFH must examine whether the tax court applied the correct method when determining the value (BFH ruling in BFHE 193, 326, BStBl II 2001, 183, under II.2para. 28; most recently BFH rulings of 29.10.2019 - IX R 38/17, BFHE 267, 18, para. 42, and IX R 39/17, BFH/NV 2020, 681, para. 40).

14. 2. the lower court only partially observed the aforementioned legal principles on the binding effect of a contractual purchase price allocation and applied them to the case in dispute. In this respect, it is not objectionable under appeal law that the tax court affirmed significant doubts about the purchase price allocation in the purchase agreement - taking into account the specific value of the object of purchase. Since the contractual purchase price allocation fundamentally fails to reflect the actual value ratios and is not economically tenable, the tax office was permitted to carry out its own allocation.

15. a) The FA and the lower court correctly assumed that there were significant doubts about the contractual allocation of the purchase price for the condominium. These already arise from the considerable deviation between the purchase price for the land (€ 20,000) stated in the purchase agreement and the standard land value (€ 77,713). The agreed purchase price share falls short of the standard land value by around 75 % and thus - irrespective of the question raised by the lower court as to what degree of deviation can still be considered insignificant (cf. Kohlhaas, Die Steuerberatung --Stbg-- 2016, 460, 462, who considers a deviation of up to 10 % to be insignificant; also Thüringer FG, judgment of 20.02.2008 - III 740/05, EFG 2008, 1140, subsequently BFH decision of 26.08.2008 - IX B 63/08, juris, legally binding)- more than just insignificant. It is true that a discrepancy to the standard land value is only a refutable indication. However, the assessment of the lower court that the plaintiff had not refuted this evidence by other evidence cannot be objected to on appeal. The tax court was unable to identify any particular aspects that would make the contractual purchase price allocation appear plausible. It rightly did not follow the plaintiff's general arguments regarding the (special) features and the renovation status of the apartment as well as the location of the property.

16. b) Since the contractual purchase price allocation thus fundamentally fails to reflect the actual value and does not appear economically tenable, the FA and the tax court were entitled to make their own allocation.

17. (3) However, the tax court was not allowed to replace the contractual purchase price allocation with the allocation determined with the help of the BMF's working aid, at least in the case in dispute. The working aid does not guarantee the allocation according to the actual market values of the land and building as required by case law. This follows from the fact that it restricts the available valuation methods to the (simplified) asset value method and bases the purchase price allocation on inadmissible parameters. The estimate of the tax court (§ 96 para. 1 sentence 1 FGO in conjunction with § 162 AO) can therefore not be upheld.

18. a) The working aid has no binding effect on the parties involved and the tax court. It is neither a legal norm nor an administrative instruction binding on the tax authorities, but - in terms of procedural law - merely a party submission by the FA. Insofar as the working aid is treated as de facto binding for the taxpayer in the practice of the tax authorities, there is no legal basis for this. It does follow the case law of the Senate on the allocation of total purchase prices insofar as it determines the value of the land and the value of the building separately (principle of individual valuation); the so-called residual value method (cf. only BFH ruling in BFHE 193, 326, BStBl II 2001, 183, under II.2., para. 29) is correctly not applied.

19. b) However, the working aid does not meet the requirements of the supreme court case law on the choice of method and could therefore not be used as a basis. It is based on the SW-RL and the NHK 2010 (Annex 1 to the SW-RL) and thus always and only refers to the asset value method; it is therefore a simplified asset value method (same opinion Schaper, Grundstücksmarkt und Grundstückswert --GuG-- 2017, 100, 101). Although the Senate has regularly considered a purchase price allocation using the asset value method to be appropriate for rented condominiums held as private assets - at least in the past (see Senate judgments in BFHE 267, 18, para. 43, and in BFH/NV 2020, 681, para. 41; from 11.02.2003 - IX R 13/00, BFH/NV 2003, 769, under II.2.a, para. 15; in BFHE 193, 326, BStBl II 2001, 183; Senate decisions in BFH/NV 2018, 218, para. 6; in BFH/NV 2017, 292, para. 5; from 23.06.2005 - IX B 117/04, BFH/NV 2005, 1813, under II.1.a, para. 3), the working aid already contradicts the previously explained equivalence of the available valuation methods from the outset, as it is not possible to assess in each individual case which valuation method appears appropriate according to the actual circumstances. The BFH has always emphasized that the decision as to which valuation method is to be used must be based on the actual circumstances of each individual case, so that the choice of valuation method cannot be generalized and in any case cannot be limited to one valuation method (cf. BFH judgments of 27.06.1995 - IX R 130/90, BFHE 178, 151, BStBl II 1996, 215, under 1.b, para. 15, and of 02.02.1990 - III R 173/86, BFHE 159, 505, BStBl II 1990, 497, under I.2.c, para. 29; BFH decisions in BFH/NV 2018, 218, para. 6, and in BFH/NV 2005, 1813, para. 1.a, para. 3). There is also no provision for checking the plausibility of the results of the working aid by means of other valuation methods, for example in accordance with the income capitalization approach widely used in real estate valuation practice. In addition, there is no justification for the choice of method in individual cases.

20. c) Furthermore, the working aid contains a systemic error in that no so-called local or regionalization factor is taken into account when determining the building value (NHK).

21. aa) The cost parameters of the NHK 2010 are federal mean values (Kleiber, Marktwertermittlung nach ImmoWertV, 8th ed. Sachwertverfahren Rz 137). When applying the working aid, for buildings of the same size, the same type and the same year of construction category, the same building costs always result nationwide due to the model, only reduced by the age value (Jacoby, Kaufpreisaufteilung für bebaute Grundstücke - Problematik und Lösungsansatz, 2018, p. 120 and 132; Jacoby/Geiling, Deutsches Steuerrecht --DStR-- 2020, 481, 483). In high-price areas, however, the actual construction costs can sometimes be considerably higher than the cost parameters (Schaper, GuG 2017, 100, 101). The lack of consideration of local conditions when determining the building value leads to a disproportionate share of the land and thus to sometimes very low building valuations, especially in large cities with high standard land values - as in the case in dispute (same opinion Rade/Stobbe in Herrmann/Heuer/Raupach, § 6 EStG Rz 315; Jacoby, loc. cit, p. 3; see also Jardin/Roscher, Die Immobilienwertermittlung aus steuerlichen Anlässen, 2019, margin no. 175 on the "damping" of the land value, for example as part of the so-called Munich model). This particularly affects high-quality properties and (renovated) old buildings (Kohlhaas, Stbg 2016, 460, 465; Jacoby, loc. cit., p. 3). This is a systemic error in the guidance, which tends to lead to an excessively high valuation of the land (same view Kaminski in Steuerberater Handbuch 2019, 27th ed., Part 4, Immobilieninvestitionen durch Privatpersonen, Rz 1273; Wagner, Der Betrieb 2016, 556, 559; BeckOK EStG/Graw, 7th ed. (01.05.2020), EStG § 7 Rz 185).

22. bb) This goes hand in hand with the fact that the building value determined in accordance with the working aid only has a limited market reference: Although the land value is determined according to the standard land values set by the locally competent expert committee and thus reflects the general value conditions on the property market within the meaning of Section 8 (2) of the German Income Tax Act, the value of the building is not always determined by the market.Although the land value is determined according to the standard land values determined by the local expert committee and thus takes into account the general value conditions on the property market within the meaning of Section 8 (2) no. 1 ImmoWertV (market adjustment), a building value that is not in line with the market is applied on the basis of THK (Kohlhaas, Stbg 2016, 460, 463; Jardin/Roscher, loc. cit, para. 176). However, land values and production costs cannot be directly compared (Jacoby, loc. cit., p. 3, 90; Jacoby/Geiling, DStR 2020, 481, 482; Kleiber, loc. cit. Sachwertverfahren para. 8). The NHK 2010 are purely model values; the market reference is only established by the asset value factor (Mann, GuG 2017, 17, 18; Schaper, GuG 2017, 100, 103; Seitz, GuG 2017, 142, 143; Jardin/Roscher, loc. cit., para. 148, 1030), which the guidance does not take into account. It is true that the tax authorities point out in the instructions to the working aid that a market adjustment of the (provisional) tangible assets has been dispensed with "as this affects the land on the one hand and the building on the other in the same proportion. The sum of the individual values determined (provisional material values) does not show the market value without market adjustment". However, when determining the land value, the required market reference - as described above - is guaranteed. Against this background, there is a deficiency in the valuation system of the working aid, which is exacerbated by the extremely dynamic development of the real estate market in recent years. This shortcoming is all the more significant as the data structure underlying the building valuation on the basis of the NHK 2010 represents the "heart" of the working aid (Jardin/Roscher, NWB - Steuer- und Wirtschaftsrecht 2014, 3155, 3160; this, loc. cit., para. 160). This does not meet the requirements of § 194 BauGB, which requires the inclusion of market conditions in the market value assessment for the entire developed property (same opinion Jacoby, loc. cit., p. 138).

23. cc) Contrary to the previous instance, this systemic error cannot be successfully countered with the argument that the "imbalance" described is a matter of imponderables that are ultimately inherent in every valuation procedure. This is because an estimate for the purposes of purchase price allocation is only to be taken into account if its basis is comprehensible and convincing (e.g. BFH ruling in BFHE 193, 326, BStBl II 2001, 183). However, precisely this is not the case with regard to the described systemic deficits of the purchase price allocation in the simplified procedure of the tax authorities.

24. dd) The lower court also recognized the "distortions" resulting from a disproportionate increase in construction prices in metropolitan areas (as in the case in dispute) - without a corresponding reflection in the working aid. However, in the absence of precise figures on the development of construction prices in the federal state in question, it did not consider it possible to intervene in the working aid to correct the situation. However, this consideration cannot be used to uphold the result of the tax authorities' estimate. Instead, the building share must be regularly determined by an independent sworn expert's report (see 4.).

25. In this context, the BMF wrongly refers to the fact that the index for building service prices of Statistical Office X has only slightly exceeded that of the Federal Statistical Office since 2015. The statistics referred to may show that the development of construction prices in the federal state in question does not differ significantly from the development of construction prices in Germany as a whole. However, this does not eliminate the deficits in the valuation system described above.

26. 4 The matter is not ready for decision. In the second instance, the tax court must make the necessary factual findings on the value of the land and buildings. In the present case of a disputed property valuation, it is generally required to obtain the expert opinion of a publicly appointed and sworn expert for the valuation of properties pursuant to Section 81 (1) FGO; it can only refrain from doing so if, exceptionally, it has the necessary expertise itself and sets this out in the grounds for the decision (cf. BFH decisions from 07.01.2015 - I B 42/13, BFH/NV 2015, 1093, para. 9; from 21.12.2011 - VIII B 88/11, BFH/NV 2012, 600, para. 4; from 03.05.2001 - III B 52/00, BFH/NV 2001, 1419, para. 4; see also Senate ruling of 06.02.2018 - IX R 14/17, BFHE 261, 20, BStBl II 2018, 522, para. 19, on the furnishing surcharge). However, an expert opinion from the tax authorities' building expert must be treated as a private expert opinion in tax court proceedings. The tax court can only base its decision on such an expert opinion if none of the parties involved raises substantiated objections to its accuracy (e.g. BFH decision in BFH/NV 2015, 1093, para. 15, with further references).

27. 5 The decision on costs is based on Section 143 (2) FGO.

https://www.bundesfinanzhof.de/de/entscheidung/entscheidungen-online/detail/STRE202010257/