A Visual Of Descriptive Statistics

Published Date: 02 Nov 2017

2.3 Results

To watch into a better way the results, it can be helpful to observe at graphs and tables [1] . In this section first of all we analyze the sample, as stand-alone populations; secondly we look at the correlations between ratios of value drivers over price (or enterprise value). After such a general overview, we will comment the results of the two empirical analyses described above using specific Inter Quartile Range and Frequency distributions for the error computed with the aim to state which are the most indicative and "best approximation" multiples in order to use them to compare the three different sectors.

Important to specify that we represent the multiples as value driver over price(or enterprise value), we invert the denominator with the numerator and vice versa with respect to the traditional multiple representation to make the analyses easier, but the meaning of the multiples, when we compare them, remain the same even if we have inverted their composition, as all of them are inverted; therefore in mathematical terms their sense is the same as if we would compute them in the traditional way, and also the conclusion can be drawn for their specific financial meanings. [2] 

2.3.1 A visual of descriptive statistics

To better analyze the samples we can take a look to the value driver over price ratio, observing descriptive statistics for each industry [3] .

As descriptive statistics we compute the Mean, the Median, the Standard deviation and the most relevant percentiles of the multiples for each sector, also values at 1% and at 99% to observe features of the ratios at the extremities of the distribution.

As regards the luxury industry here we have no negative values between the relative means neither between the Medians. If we contrast mean and median we see that fcf/p, sales/p and the two forward looking ratios have the most symmetric distributions. Standard deviation is really low for the most of values, meaning that the observations are really close to their averages. The square root of the variance find a higher value only for operative cash flow over price ratio, as cfos usually have higher numbers compared to the other value driver; also if we contrast them to the free cash flow, probably for the inclusion of the capital expenditure in their computation, those ratios have higher numerators than free cash flow ones.

As regards the first and the third quartile, other than for the cfo/p, they seem to measure a quite normal spread of the numbers in the interquartile range, exception for sales/p for what we find that the major part of its observations are more concentrated before the third quartile; also looking at its interquartile range it is quite evident that between the 25th and the 75th percentiles observations are more frequent.

Concerning the most extreme percentiles the lowest 1% pertains to cfo/p and is negative; but having the highest standard deviation this ratio can be considered as an outlier compared to the others. We also find that the lowest first percentile belongs to fcf/p, which means that cash flows measures have most observations of the lower numbers. Looking at the other extremity, other than for the operative cashflow over price, we find higher values for the 99th percentile in book-value/price, sales/p ratio and, for the same value numerator, also sales/ev.

Analyzing the automotive sector we have one negative value under the Mean column, belonging to the Earnings per share over price ratio, and comparing the Means with the corresponding Medians we find that only the book value/price has a symmetric distribution. Cfo/p has the highest variance in the sector, followed by eps/p having a SD of 14, 84; higher variances, if compared to the general SD level in the sector’s sample, belong to sales/price and to book-value/price. Ebitda/ev is the ratio having values less far from the mean and it can be justified by the fact that this ratio is not affected by the capital structure of the company [4] ; in other terms, even if in time of crisis a cyclic sector as the automotive tends to increase its level of debt, the ebitda/ev will not change significantly.

The interquartile range of the motor vehicles industry, get the highest value for the cfo/p ratio, followed by the sales/p ratio; in fact the sales over price ratios have a small number of observations after the third quartile.

Regarding the other percentiles, the first measured (1 %) has negative values for cfo/p, fcf/p, eps/p, adjusted eps/p and book-value/price. The maximum percentile (99%) finds the highest values in the ratios having higher standard deviations, in particular cfo/p, sales/p and eps/p. Looking at the other percentiles we note that fcf/p keeps negative values until the first quartile, for the automotive sector, showing that some companies in the sample are quite wasteful, for the part considered in the sample. Moreover also the adjusted Eps over price has a negative value, but only until the 5th percentile, as we can see in the descriptive statistics for this industry.

Concerning the counter cyclical food industry, if we put in contrast the values of mean and median, we find that fcf/p, eps/p and ebitda/ev have the most symmetric distributions; the less symmetrical one belongs to cfo/p. As regards the Standard deviation, between the staple food sector’s ratios, it has the highest value for cfo/p; after, we have a high variance in the sales over price. Also here the lowest standard deviation value pertains to ebitda/ev, for the same reason explained before in the automotive case.

As before the cfo/p has the largest interquartile range. Also sales/price and sales/ev ratios have most of their observations before 1, 25 and 0, 94 respectively.

As regards the first and the last percentiles computed: we find that the lowest 1% are the book-value/p and the fcf/p one, instead the highest 1% pertains to cfo/p, as for the previous sectors; as regards the last percentile computed it finds its highest value in the cfo/p, followed by the two sales ratios, as anticipated by their standard deviations.

Therefore, from a visual inspection to the descriptive statistics of the Staple Food sector, we have observed that it has no negative value even if we are considering also data during years of the crisis; these results can be seen as a confirmation of the staple food counter cyclical nature.

2.3.2 Correlations between different ratios per sector

It is quite interesting to take a look to correlations between different ratios per each industry to see if they can be explained by the same variables, as their value drivers can have some observation in common, included in their calculation [5] .

Let us start with cash flow measures, operating cash flow over price and free cash flow over price have been calculated with a similar formula: the only difference is the capital expenditure which is excluded to compute the fcf from the cfo. The correlation is around the 80% for the luxury and the staple sector.

Continuing with the free cash flow, we can see that it has positive correlations with the other ratios (as sales/p,ebitda/p, sales/ev, ebitda/ev) only for the staple food industry, as far as ebitda/ev has a negative correlation for the automotive; this negative result can be justified by the ebitda/ev neutral capital structure nature. We find another negative correlation between fcf/p and eps/p in the motor vehicles industry, but it has no significance if we look at its corresponding p-value level.

Looking at the Cfo/p correlation with different ratios, it shows a high positive correlation with sales ratios, ebitda ratios, eps and book value for both the luxury and the staple food sector, slightly higher for the last one. There results can be explained by the cfo/p meaning which is clearly related to the other ratio numerators, as it states how much cash the firm can generate from its working business.

Sales/p has a strong positive correlation with Ebitda/p and sales/ev for all the sectors: with the latter as they have the same numerator, with the former as ebitda is computed starting from sales revenue.

Is interesting to study the results coming from sales/p with eps/p, as what come out are a negative value for the automotive and a significant positive one for staple food. Multiples theory explain that price/sales multiple indicates how much we have to spend to obtain an amount of revenue per share. Instead p/eps states how much we have to invest to obtain earnings per share. Comparing these two categories of multiples we can conjecture that in a cyclic sector (as the automotive one) as a crisis has just occurred, the real worst results will come the year after creating such a no sense between two positive correlated value driver: earnings and sales [6] .

The remaining ratios are always positively correlated for both the luxury and the staple food sectors, only a few of them have a p-value greater than alfa, and they are all correlations between multiples of the automotive sectors.

Attractive are the cases when we find positive correlations between ratios for all the sectors: Ebitda/p with sales/ev for example. When we have defined the relative multiples [7] , we saw that those two have a quite different meaning as also because value driver are computing over two different bases: price and enterprise value. But we also said that when we use enterprise value it can be considered as an extension of the market cap, as it has a positive relation with the market value of a company. Also we found, explain other correlations before, that ebitda comes from sales revenue. Being positively correlated both value drivers and both price and enterprise value we find a high level of correlation for all the industries considered in the analysis.

The multiples of the automotive sector show not so strong correlations also in the other comparisons, sometimes even negative, but only slightly (values greater than -0, 5), the most of them have no significance.

As regards correlations between forward-looking value drivers over price (or enterprise value), they also show positive strong correlations between multiples in the luxury and in the staple food industry; the lower correlation between them in the automotive industry has no significance.

2.3.3 Results of the traditional valuation

From the traditional valuation multiples, under the hypotheses of direct proportionality, we have computed the Betas for each year in three different sectors. In other terms, in this first stage of the analysis we have as dependent variable the price of a firm and as explicative variable one value driver; what we want to state here is that there exist direct proportionality between the value driver and the market price of a firm. Using those Betas, computed using the harmonic mean of value driver over price of the ten firms per each sector we compute the pricing errors scaled by the relative price. In statistic the error states how much the sample deflects from the expected value (supposed to be equal to the unknown real value). For the previous hypotheses, we should obtain independent errors and we have to take a look at how much their frequency distributions approximate a normal one, to see which has a lower dispersion (between the ratios that we are analyzing).

To study the errors we compute the interquartile range of their distribution, for each sector separately, across the twelve years [8] . As regards the luxury industry we find that here the errors are independent as we can see from the graph they do not show a common trend or also a trend, if we watch them as stand-alone; the same conclusion for the staple food sector.

A different course is followed by the interquartile range analysis in the automotive sector, but going deeply in the ratios data, used to compute them, we find that this behavior is mainly caused by the free cash flows IQR line. From the free cash flow value driver it comes out a Beta, in 2002, which is too high (compared to the rest of the sector in the other years considered), as during that year the firms have quite similar free cash flow values, compared to the other year in the sample. Particular events in the sector happened during the following years make the free cash flow of the sector more different and more scattered from each other [9] . Leaving this "outlier" multiple we can see that the interquartile range representation for the pricing error for the sector is similar to the one for the other two industries. We can justify this uncommon behavior of the value driver with the fact that this cash flow measure is computed excluding the capital expenditure from operating cash flow, where expenses for the capital are important elements mostly in the automotive sector; in fact in this field physical assets are more necessary than in the other sector studied, that’s why we find such a large difference between cfo and fcf trend in this specific sample.

Continuing on the view of the errors, we compute frequencies for each multiple for each sector on the same line of bins going from a value of -2, 1 to 1, with a width on the x axis equal to 0, 1 [10] .

As regards the luxury industry from the charts we can observe that most of the errors approximate a normal distribution; in particular the best ones are the ebitda/p, ebitda/ev and sales/ev.

The automotive industry shows a different pattern of its errors distribution. Taking them separately we can see that similar to the shape of the normal distribution we have sales/p, ebitda/p, sales/ev, ebitda/ev and book-value/p. Instead the two historical ratios pricing errors farthest from the normal distribution shape are fcf/p and eps/p, which seem to have quite similar frequencies over the same bins. Earnings per share can help us to state how profitable is a firm, especially from the shareholder point of view; looking at relations and implications in financial statements we find that free cash flows, being the rest of cash that a company produce after the deduction of capex, have an important effect on eps. In particular, past studies note that if fcf rise today, "tomorrow" this enhancement will influence eps in a positive way. In other words eps growth is supported by the increasing trend of the company’s fcf. This relationship can explain why they have similar frequency distributions. [11] 

As regards the counter cyclical staple food industry, excluding book-value/p and eps adj/p, all the other ratios pricing errors have shapes that approximate a normal distribution, as we have seen for the luxury sector. As regards the bv/p errors frequency, going back to the ratios we see that book value has not a serious drop, instead the sample show a general positive trend, where the best in the sectors is danone(in liquidation value terms). The problem come from the market price, as the drop in 2008, when the crisis started, was not compensated by the value drivers for this ratio, as the book value is a more stable measure of the company [12] , that is why as a strong event occur the more sensitive variable is the market price which changes the ratio.

2.3.4 Results of the valuation with an intercept

In the second stage of the analysis we have abandoned the direct proportionality assumption between value drivers and prices computing an intercept alfa. Doing that we have included in the calculation of the price some omitted coefficients improving the whole analysis. In other words we are supposing that there are other variables, other than the value drivers, which influence the market price of a firm; what we do in this stage is to see if including these variables the dependent one (the price) will be explained better. To state the last finding we have to observe also here the independence of the errors and for which multiples’ errors the frequency distribution is less dispersed.

Also here, as in the previous stage, we study the distribution of the pricing errors. Starting from the interquartile range we have computed it for each multiple for all the sectors taking their pricing errors as a whole. From the graph [13] we can see that pricing errors are independent across the twelve years.

Studying the frequency distribution of the pricing errors [14] here we find curves which better approximate the shape of the normal distribution. They are all skewed on the left, which is caused by the presence of small outliers. The distribution are computed for each specific industry ratio’s pricing error resulted by the second stage analysis.

As regards the luxury industry we can see that the best approximating distribution pertains to ebitda/ev. Instead the ratio’s pricing error having a shape which is really distant to the normal one is the distribution of the sales/p. The problem is that the sales ratio is limited to the story of the firm, it can be helpful, sometimes, only to compare companies belonging to the same sector, but often is meaningless also in such a comparison as it does not take into account, in its computation, many important variables (as the debt for example) [15] .

Looking at the automotive industry distributions, we find that almost all the ratios distributions are more skewed on the left (if matched with the relative ratio’s pricing error distributions of the other two sectors). The chart most similar to a normal distribution is the ebitda/ev one, as for the luxury industry, even if the automotive shape one is slightly more skewed than the luxury one; also for this sector the more skewed distribution is the one of the sales/p ratio.

The staple food sector, instead, seems to approximate the normal distribution more in the cash flows ratios pricing errors. Also here the ebitda/ev has a good distribution of its errors, but is more skewed if compared with the last two multiples. The worst distribution for the food sector belongs to sales/p pricing errors, as for the other two sectors.

We also compute the frequency distributions taking pricing errors together from the three sectors per each multiple on the same bins used in the previous section (from -2,1 to 1, with a width of 0,1) [16] . Here we obtain one frequency distribution per each multiple, what we find is that they approximate, more than in the traditional analysis and more than the last one (where we have considered frequency distributions multiple in each field), the normal distribution shape. In particular, compared with the normal one, these distributions are more skewed on the left tail, which means that median values are larger than mean ones. Negative skewness comes from the fact that there are small outliers values in the sample considered. Observing the charts we can note that the less negative skewed distributions are the ones of the enterprise value ratios (sales/ev and ebitda/ev). This phenomenon can be explained by the fact that, as we have said previously, the enterprise value is an enlargement of the market cap value of a company, in which we include also its market value of debt; this measure is a relevant factor in times of crisis as it influence a lot the value drivers of a company, and also because it has a different role, during a recession, in cyclical rather than in counter-cyclical sector.

2.4 Conclusions: is the luxury industry counter-cyclical?

Looking separately to the ratio distributions per each industry, we have to keep into account that exist multiples more adapted to each sector. Analysts have studied which are the most indicative ones studying the variability of each value driver in the financial statements of the companies. From the paper "how the analysts reach their conclusions" of Pablo Fernàndez we see, among the ones calculated in this final work, the financial ratio specific for each industry. As regards the luxury sector, the most used multiples are: Enterprise value over sales and enterprise value over ebitda. Concerning the automotive we find the Price to sales. Finally in the staple food the most used is the enterprise value over ebitda.

When we have discussed the results of the two valuations, we saw that in the first analysis the less dispersed pricing errors in all the industries are the enterprise value ratios (ebitda/ev and sales/ev) and the ebitda/p; in the second one, in which we have included other factors allowing for an intercept, we found that ebitda/p and ebitda/ev have the best approximation of pricing errors distribution to a normal one. Having proven that these last ones are the best in the valuation of the price, we can use the correlations between the sectors under these ratios, extrapolating them from the following correlation table, to help us to compare the three industries.

correlation between

luxury-auto

luxury-staples

FCF/P

0,58430594

0,770653034

CFO/P

0,0267346

0,785238307

SALES/P

0,851916276

0,603016665

EBITDA/P

0,745042346

0,818501313

SALES/EV

0,412127495

0,560193981

EBITDA/EV

-0,107399572

0,556099311

E/Price

-0,718507183

0,590041978

BV/P

0,355964986

0,639101883

Last table consists in the study of the correlation of the same ratios between different industries. We can describe the correlations for the most guidance ratios for each industry.

Concerning Sales/p, here the correlation is highly significant more between the cyclical industry and the luxury (0, 85) than between luxury and staples (0,603). But here we have to consider that this ratio, even if it is the most indicative for the automotive sector, is almost meaningless if used to compare such three sectors, as it explains only the story of the firms, and does not include in its calculation expenses neither debt’s value. So we can ignore these two correlations.

As regards the sales/ev, which is one of the most indicative ratio for the luxury industry, the correlation between this sector’s ratio and the staples one is higher than 0, 5 (specifically 0, 56).

The ebitda/ev, which is indicative for both the luxury and the staple, has a coefficient of 0, 55 between those two sectors.

After this analysis we can see that, for their most indicative multiples, the sector which is counter cyclical by definition (the staple food) and the luxury are highly correlated.

Having summarized the relationship between industries with the best guidance ratios, after having stated that use these last ones we can measure the equity values of the companies, we can see which are the implications of the results for the luxury nature; in other words we can define if it has a cyclical or counter-cyclical nature during the actual financial crisis.

Before the final conclusions, we have to note that the core industry of this final work has been affected by the financial crisis the year after the start of the recession (2009); we have seen the same trend for the staple food industry, which is counter cyclical by definition. It means that the crisis has affected also counter cyclical sectors, but in a lighter way compared to the performance of a cyclical sector as the automotive. In fact empirical researches show that staples have suffered the recession because people continue to buy them daily, but lower quantities of them have been sold during the worst year of the financial turmoil; this drop has been also caused by the policy of no- waste which has been spread all over the developed countries affected by the crisis. The same can be said for the luxury industry, as here people have bought less precious items during the 2009, and also its companies have suffered the economic crisis as all the other sectors, but only for that year.

After we have seen the meanings and the relationships of the different multiples for each sector and the trends of the three industries, we can state that the luxury industry was and is being counter cyclical during the actual crisis.

THIRD PART: THE LUXURY FINE WATCHES

Luxury is strictly linked to scarcity and to the adjective "advisable". As we have seen in the first part of this work, this industry finds its sense in people who want to spend a lot of money to fulfill their desires or to buy something only because it is difficult to obtain. Across decades the definition of this sector has changed in such a way that nowadays we have a concept of luxury completely heterogeneous. Talking about heterogeneity, it’s really important to highlight the existence of many aspects of the luxury industry, in particular on those ones which come out if we classify different segments in terms of shortage and materiality of the specific goods. Through this analysis we find trends called niche luxury, connoisseur luxury and super-luxury.

Empirical studies have showed that there was a dominant trend in the luxury accessories segment already in 2007: the growth of the industry was mainly led by the super-luxury brands. In particular one interesting trend is the one of the luxury watches, which seemed to be really upward during last years (with the exception of the 2009). It has been also said that the luxury fine watches segment can be classified as the most profitable segment of the luxury industry.

3.1The boom in the luxury watches market

Notwithstanding the recession, some exclusive precious luxury items are still in high demand. Remarkable is the case of the so-called "hard luxury" market, where for hard we mean unique goods as jewels and timepieces.

The boost of the hard luxury brands, during the last 10 years, can be explained by many points of view. First of all there has been a mix of marketing strategies: first and foremost we find the premiumization, and after that the existence of more consumers which are mostly attracted by goods with value-added characteristics, compared with the past; the last but not least has been more noticeable for some luxury watches brands, as for example Richemont. Secondly, a great contribution to the upward trend of the hard luxury market come from the geographical diversification oriented toward higher-growth countries, as China for example; in particular Asia and Middle-East and Africa have seen a growth in their stake in the jewelry market of about 35 %. Third point, the boom of the hard luxury has been caused by a demographic strategy geared towards advantageous consumers, both for demographic age and for gender-taste consumption differentiation.

Innovative events and already existed fundamental aspects together generate good expectation for the "Jewels-Fine watches" market. For instance Swatch Group protects the Swiss tradition keeping as supplier Eta, which produce the most of Swiss timepieces from years; as regards leading jewels protagonists, Richemont remains the largest with Cartier. But what the market can forecast for these firms can’t be said for big conglomerates as LVMH or PPR Group, as here what may be relevant consists in Merger & Acquisition business (as we have seen in the past when Bulgari was acquired).

Finally in the following chart it is possible to observe the trend of the luxury timepieces world in terms of market size, where the axis of the ordinates has as unit of measure millions of $.

3.1.1 Luxury watches market history: Made in Switzerland [17] .

When someone talks about fine watches is implicit the Swiss manufacturing and technology which characterize them. The entire luxury watches market comes from Switzerland, and across the previous century it has been spread all over the world.

Most relevant in the history of this market has been the "quartz revolution" which led almost ten years of stagnation for the segment. The growth restart during the 80’s, ever since it continues for the last 27 years keeping an amazing upward trend also as the current crisis was beginning: in terms of CAGR mechanical innovations led it to a growth rate of more or less 5% in the exportation of Swiss timepieces during this development time horizon.

What nowadays still identifies the uniqueness of a fine watch is the manufacturing of the basic movements, the so-called "tractors" whose efficient production requires years of experimentation and high levels of funding. One of the most representative brands using this mechanism is Swatch. Swatch was and is investing a lot of money in the basic movements, guaranteeing also an easy repair service to customers who purchase its timepieces. It has been said that this could be the investment allowing the leadership position of Swatch Group in the sector. At the same time other luxury fine watches brands offer exclusive and unique products because of specific mechanisms created and secretly retained in their companies through patent’s issues, as for Cartier. Moreover it also happens that big brands, even conglomerates sometimes, use to pay millions of CHF to old craftsmen/watch cases of the place to purchase their "secret mechanism": it has been the case of Bulgari with Finger.

The Swatch group success is also caused by the fact that its "basic movements" has been used by many others. Maybe for its lower price of investment or for its enormous standardization, the swatch mechanism has been adopted in three different ways: entirely took over without permission of the original maker, or adapted to the specific mechanism of the other company, or entirely used mentioning its origin. And it has been for the first mentioned kind of user that the Swiss group closed the supply of the "tractors" to the other companies few years ago.

Recently also Chinese mechanisms for watches, in terms of manufacturing, are trying to reach the top of the market and to become real competitors of Swiss producers; but the latter are still beating the former in terms of quality.

3.1.2 A different story of success: Compagnie financiere Richemont AG

The Swatch "basic movements" and the pure quality of watches made in Switzerland seem to fade the hope that an excellent watch firm could come from other place, using other business strategy to reach the success. That is why is interesting to look at the story of a company which was born originally from a spinoff and consequently had made several acquisitions: the Compagnie Financiere Richemont AG. In its full, even if short, journey, the most impressive part of the M&A activity has been the acquisition of Jaeger Le Coultre, of A Lange & Söhne and of the 100% of IWC during December 2000. In this way, thanks to the power of synergies coming from merger and acquisition activities, the company obtained original skills, both in terms of watch production and geography, putting down new roots in a different place compared to its origin: the timepieces motherland.

With more than 1 Billion of Euros, during the first year of the financial crisis, Richemont has been defined as the "largest hard-luxury player".

Luxury watches contributes to the growth of the company for the 50%, and is one of the most significant element of the group, not only for the percentage of profit generated by its sales, but also for the broad span of customers achieved with its exclusive timepieces.

Therefore Richemont occupies a large stake of the luxury watch market; most important brands through which the company covers the highest segment of the luxury industry are A. Lange & Söhne, Piaget and Vacheron Constantin. It operates also through Montblanc and Cartier, being those two ones more skilled and experienced in the manufacturing of other expensive masterpieces.

3.1.3 Asian scenario

"Riding the gilded tiger" says the headline of an article in The Economist of May 2012 [18] .

In particular the article contains opinions regarding the forecasts of the annual growth in luxury goods, and it says that for more than half of 2012 growth came from China.

Analyzing market sizes forecasts for the next 5 years [19] , it is quite obvious that across the time China is overtaking Japan and becoming the second largest luxury market in the world, after the Us. China will grow with a rate of 15, 4% between 2013- 2014 and the 13, 3% between 2016- 2017, according to the provisions.

The total consumption of luxury industry products has been almost equal to 60 billion of euros in this country, during the last year, and this is why China can be also defined as "the country with the greatest purchasing power in the luxury".

Continuing on the topic of the luxury watches market, in terms of international auctions, which are one of the basic channels to distribute such masterpieces as they can also be considered as collectible artworks, we find three main actors in this auctions’ market: Switzerland, United States and China. Geneva is maintaining its leadership status from 2010 to 2012, with a portion in the international turnover, in the first semester of the last year, of more than 42%. Immediately after the motherland of timepieces we find Hong Kong, which in terms of international turnover has held the 33, 6% in the first six months of the last year, and in terms of volumes sold during the international auctions has outperformed the other two actors. As regards the United States, New York resized its portion in the auctions’ total turnover during 2011, and last year was recovering share even if is crystal clear enough that this actor has been beaten by China, as it can be seen in many other facets of the luxury market. Asia is the strongest actor if we consider the timepieces auctions’ unsold rates during the last five years.

The reality is that the luxury watches are challenging the economic crisis, going against negative trends with positive results every month. As regards the Swiss export, also here Asia is the main pulling, even if China was not keeping up by itself the pace during 2012.

Notwithstanding 2012 Chinese shortcomings, Hong Kong plays a key role in the Helvetic exportation of luxury timepieces keeping the leadership in the market, obviously followed by the United States. After them we find Japan, recovering its position compared to the past years, and Singapore keeps a good ranking even having low growth rate in the Swiss watches export.

One practical example or case study can be the one of Richemont in Asia, as the company has already been established in the continent, now it is also in the Chinese hinterland. After investments in infrastructures and training schools, Richemont is still improving its distribution network. But it is not the only one taking advantage of the growing Chinese luxury market: Swatch Group closed the 2012 with almost 6, 7 Billion of euros of revenues thanks to sales of Watches and Jewels in Asia, as expected by the investors and as happened in 2011 when the Asian customers (because of the East continent economic boom) increase the demand for the Swiss company’s timepieces.

3.2 Luxury watches indices

As in the first part of this final work, we can build stock indices to observe the performance of the sector across time. As before, we use two methodologies to construct indices: price-weighted and market-cap weighted. Let us see how to construct the sample.

3.2.1The sample

It has been said that the motherland of the luxury fine watches is the Switzerland, and that in terms of auction turnover, alternative basic mechanism production and exportation China has a primary role in the timepieces industry, especially during the last years, when the Asian boom has been started. Thanks to the Asian boom also Japan has regained ground and, lastly, the United States still cover more than half of the total turnover at the international auctions of fine watches.

Having said that we decide to construct our index taking the most famous timepieces firms quoted on the respective country financial markets, as the aim is to see the behavior of this good internationally, and the best way to see that is to observe stock exchange quotations.

Starting from the motherland, we take two firms quoted on the Swiss Stock Exchange, the most famous we have talked exhaustively about so far: Swatch Group and Compagnie Financiere Richemont. Then we choose to include Emperor Watch and Jewelry listed in the Hong Kong Stock Exchange, a firm founded in 1942 in China, where Emperor is synonym of excellence. From Japan we have three firms: Rhythm Watch Co, Seiko Epson Corp and Citizen Holdings Co Ltd. Japan luxury watches are much more treated in their technology than in the aesthetic; the advanced mechanisms and particular material such as silicone are being used to make the timepieces unique and more difficult to build, compared to the other ones, and therefore less affordable to all. Last element of the sample is Fossil Inc which is an American company producing and distributing primarily watches and jewelry also for other famous luxury brands. The success of the company comes from previous acquisitions of two Swiss brands (the most important was Zodiac) which allow the firm to replicate the motherland’s secrets.

3.2.2 The timepieces stock indices

Once the sample has been completed we can start to compute the indices. The price- weighted index, calculated using the same formula in the first part of the work, has the following shape:

As we can observe in the graph, during the first four years the index underperformed, compared to the following years, because of the survivorship bias, as half of the sample was still not listed. From the 2007 there has been an upturn of the luxury watches price weighted index followed by a drop during 2009, year of recession for the entire luxury sector. Thereafter the index remains above layers of at least 500 $.

To compute the market-capitalization weighted index for timepieces, we use the Paasche pondering in the formula. [20] What we obtain is the black line in the following chart, which from July 2010 has a tremendous upturn. As benchmark we took four value-weighted indices, in particular we choose to take four indices corresponded to the countries of provenience of the firms in the sample: S&P 500 for the United States, Swiss market index for the Switzerland, Nikkei 225 for Japan and the Hang Seng Index for Hong Kong.

As it has been shown in the chart, the timepieces market capitalization weighted index, after the period of recession (2009), has successfully outperformed the other market indices growing by 87 %; so the luxury watches sector can be considered as a "safe haven" in which invest during this period of financial crisis.

In particular, the benchmarks used in the chart have had the following percentages of growth from July 2008 until now:

mkt-cap weighted lux watches

SPX Index

SMI Index

NKY Index

HSI Index

% of growth

87%

14%

13%

-1%

4%

How can we explain this huge difference between the benchmarks and our index?

The upward trend of the market cap fine watches index has not been affected by the volatility that has marked the other industries. Thanks to what we have seen until here about this sector, we can say that its course reacts contrarily to the "global volatility" because of the successful results of the industry’ s international auctions: increase of the total turnover and growing number of investors in the sector.

It seems like our index, if viewed as a global representation of the entire luxury timepieces sector, is not influenced by marketplace’s uncertainty coming from the actual financial situation, excluding the year 2009.

3.3 Luxury watches as underlying of an investment fund

Luxury watches can be considered as collectibles, as we have seen before. [21] Collect meaning can be expressed as the strong desire to have something, and rarer and particular the objects are, more the collector feels satisfied. In this last definition we find some words we have already heard, words that are at the base of the concept of luxury. Putting together the passion in collecting and the luxury watches, it comes out the story of Alfredo Paramico.

Mr Paramico is an Italian financial manager who spends almost the 89% of its total salary from his past career years, to buy precious watches to collect them. What he prefers is just the best for its personal collection: he buys watches that are not easy to find and qualified as higher grade ones; in other terms he mostly buys Patek Philippe. This brand is one of the most famous among the Swiss craftsmanship of timepieces. Patek produces masterpieces which are so precious not only for their appearances (as they are usually plated in white gold and platinum), but also for the mechanism as this brand use special movements and other techniques to make its pieces unique. Another important feature, as requirement for watches to be interesting for Paramico, is that they must be vintage pieces, something that tells a story, not something new. His collection actually worth 27, 3 Million of Us dollar, and it counts for a total of 10 "works of art".

The most curious part of the story starts in 2010, when an investment company working with Jewelry and high-quality wines, based in Luxemburg, proposed him to make with his collection an innovative, not already existed, financial instrument: an open-ended timepieces fund.

The Luxemburg company is Elite Advisers which has been founded in 2007 with the aim to create a new kind of asset class composed by investments in tangible goods. For Elite Advisers have a portfolio allocation in equities and bonds is superficial, and it has more volatility if compared to their innovative investment framework. Therefore they have three funds: Jewels, rare wines and, from the end of 2010, fine watches.

Elite Advisers and Paramico have created the Precious Timepieces Fund which has been launched in October 2010, and its inception date took place in March 2011. The fund invests for two third in vintage timepieces and for one third in modern watches. The higher percentages of the portfolio allocation are in Rolex and Patek Philippe, followed by other famous brands such as Cartier, Vacheron Constantin, Audemars Piguet, Longines, Omega and so on.

Paramico is The Fund Manager and, thanks to his knowledge and passion for rare watches, the fund is specializing in the purchase of the most precious pieces existed in the world. Being an open-ended fund investors have not the problem of duration’s boundary and it is possible to invest there in many ways, not only directly but also through an insurance company.

Being the fund allocation in rare tangible assets, as watches are, it can be used for defensive strategic investment so as to create an active allocation.

This kind of investment offer high perspective of long-term return, a yearly percentage of 15.

Looking at the numbers we can see that the actual Asset Under Management is 22, 83 Million of Euros, the Net Asset Value amounts to 123, 54 Euro and the 1 year performance is of 11, 66 %.

Therefore investment in such a fund is quite attractive, also considering that timepieces like those in the portfolio allocation are becoming rarer as time goes by and as they can be defined as safe havens; but it would be worth to highlight that such a peculiar form of investment could be most suitable for investors to have a deep knowledge of the field.

CONCLUSIONS

We can close this last part of the final work reviewing our main question: "Was the luxury industry affected by the financial crisis?". At the end of this work we are able to give an answer to it resuming what we have stated at the end of the three parts. In the first part we found that the overall luxury market Ipo has seen a good trend during the main years of the crisis. In the second part, after having stated that the multiples are a good evaluation of the equity, we found a strong correlation between the luxury and the counter cyclical by definition staple food industry; therefore during such years of disarray the core sector of this final work seems to be counter-cyclical to the crisis. Finally, studying the most gainful subsector of the fine watches we have stated that is a place to find safer alternative investment opportunity. To give a best view of this subsector we study the case of a fund investing in tangible assets, specifically in fine watches.

Therefore we find that the overall luxury industry does not show the same negative trend of many other sectors, both as regards the Ipo market during the last years, and also for the already publicly traded firms during the past 12 years. Notwithstanding the positive performance of the core sector, we have also seen that there has been a drop during the past years: in 2009; but then we have also noted that during this year the crisis has affected also the most counter cyclical sectors (as the staple food one).

Thereafter the luxury sector has been affected by the financial crisis, but only briefly and partially, during the rest of the time this industry continues to trace its road to success. In particular there is the subsector of the luxury watches in this industry that can be defined as a safe haven, being fine watches rare masterpieces of a quite treated manufacturing; hence fine watches stocks or even the intangible asset self, can be used to put in place defensive strategies of investment.