Since the market crash in 2008, the term “bubble”, applied in an economic context, has become part of everyday vocabulary. The analogy of a market growing so big, like a bubble, filled with air only waiting to burst, is almost irresistibly charming. Economic bubbles happen all the time, all over the world.

The most discussed possible bubble in recent years has been the art market. Critics, ranging from economists, journalists to opinionated folk, claim that the art market is a bubble and that when it bursts, it will be ugly. Maybe this talk started exactly the same day the biggest bubble in almost a century started bursting: 15 September 2008. On this day, one of the biggest Wall Street firms, Lehmann Brothers, went bankrupt, signalling the real depth of the crisis the world was about to face. On the other side of the Atlantic, Sotheby’s held a one-man auction, with works from Damien Hirst only commissioned directly from the artist, raising £111 million. This is an almost unbelievable coincidence: one of the most controversial artists selling stuffed sharks in formaldehyde for millions, making a fortune on the same day that many others lost their savings.

Since that day, all sorts of records for the most expensive work sold on auction have been set, ranging from most expensive artwork sold by a living artist to Chinese antiques. There is no doubt that the art market has seen a boom in the last decade. Prices have surged while tales of paintings being resold for many times the initial value in the space of just a few years have made front-page news, with the art market in the spotlight.

Despite media attention and the worsening of the overall world economy, the art market kept strong results going in the 2010’s, until the middle of 2015, when auction results started showing signs of weakness. Records were still being set, but lots exceeding their estimates became rarer and the buy-in rate, works that failed to reach the reserve price and came back to the sellers, went up. The hunt for prime oeuvres didn’t change much, but the pieces slightly below this level started to see more constricted bidding. Sotheby’s last quarter of 2015 showed a loss in auction revenue, leaving the full year results undershooting analysts’ expectations. The company is expecting a drop of 33% in revenues in the first quarter of 2016, signalling a belief that the art market is not so strong.

If auctions aren’t yielding the expected results and art prices are stagnant overall, are we seeing the beginning of the long-awaited art-market bubble burst?

Why people think the art market is a bubble

Media channels have been advertising the art market bubble with a certain euphoria. In 2009 BBC produced a documentary called “The Great Contemporary Art Bubble”, an extremely biased hour-long study in which dealers, artists and gallery owners predicted the imminent crash of the art market. 6 strong art years went by, with the art market growing at an average of 13% a year according to a Citi report last year, using Artnet-compiled data.

According to the study, the growth of the art market since 2000 has not been uniform on all price ranges for artworks. In what is called a “fat tale” distribution, the growth of the top 20% of the most expensive works was faster, compared to the rest. This doesn’t mean that medium-market works lost value, just that the presence of blue-chip works has grown more than others in the last 15 years. This is one of the reasons for growing media coverage of the events, spurring people to believe that these are only speculative prices. The only group able to buy these works, the world’s 0.001% richest, has grown a great deal in the last years despite overall economic depression, causing the phenomenon.

Last week Bloomberg anchor Matt Miller interviewed an analyst on the state of the art market and the last Sotheby’s results. In his concluding remarks, Mr Miller, who was unaware of who Willem de Kooning was previous to the interview, asked how the art market can fail to be a bubble if Ken Griffin could have bought his own building with half a billion dollars, instead of canvases by Pollock and de Kooning. The fact that Bloomberg journalists, theoretically educated in finance, can make this type of remark helps to show how overall opinion might be based on refutable common-sense evidence. Fuel is added to the discussion by the general public perception of buying art as a tax-evasion tool, combined with media attention and the sale of very expensive artworks by artists barely known to the population, like a Barnett Newman going for 80 million dollars.

The existence of a speculative bubble in the art market would suggest that overall art valuations are wrong and that when people realise this and adjust to the real value, a big downfall in prices will occur. The problem with a speculative bubble is that all art valuations are speculative in essence. When you are analysing a company, a bond or even oil prices, you have an underlying asset to dissect based on its intrinsic value but painted oil canvases aren’t worth much per se. Art prices mostly rely on a benchmark base, using comparisons with similar ones sold previously, generating a self-support mechanism of price setting. This mechanism is being pointed out as the bubble maker, but if things always worked this way, then why is there a problem now?

Bubbles can have different origins and reasons. The 2008 American real-estate bubble was created by excessive leveraging in financial products related to mortgages and the wrongful assumption that house prices would always go up. The dot-com bubble in the 2000s happened because tech companies had valuations based on wrong assumptions, such as the number of visits per day on a page. But one of the things that people tend to forget is that almost no one sees the bubbles before they burst. Only a handful of hedge funds made (a lot of) money in 2008, while the majority lost aggressively. Since then, people have been looking out for the big bubble scoop, in the bid to be the one who predicts correctly when and where the next one will blow. It’s not a simple task.

The “Is There a Bubble on the Art Market?” study

Last year, a study paper using econometric models to establish if there is a bubble in the art market, published by economists from the University of Luxembourg, came to public attention. They concluded that there is indeed a bubble, especially in the contemporary market. Naturally the study resonated in major art media outlets and helped to support bubble believers, but one must take a minute to understand what the authors did and what it means. The conclusions of economic papers are often – as in this case – supported by econometric regressions. How you make the regressions is the hard part since one little mistake or difference of opinion can bias the results. Data sets are also not also to be handled lightly and some of them aren’t suitable to certain types of regressions.

The dirty little secret of all art price-related studies is that the data used to build a price curve is always biased and insufficient. Artworks aren’t traded enough to establish a pattern over time and many artists who lose value just disappear from the curve because they have stopped being traded, in what is called survivorship bias. They also rely on auction data but most transactions in the art market are made through dealers and galleries. The reader must bear this in mind when discussing economic papers comparing either art prices against other assets, art returns over time, or the possible existence of a bubble.

What the study actually does is to analyse the data and test it against a model believed to show the existence of a bubble. To sum up the model in very few words, it says that a bubble exists if the behaviour of intrinsic prices for an asset does not match the behaviour of market prices over a period of years. The problem is, as previously mentioned, the analysed market prices for artworks do not represent the full reality and artworks have no intrinsic value for comparison. The authors build a model on the basis of auction houses, sales dates, media, work size, whether a work is signed and whether the artist is alive to determine the intrinsic value of an artwork. It’s safe to say that these 6 variables alone can hardly determine the value of an artwork. Therefore, the results of the regression and the conclusions for the study must be treated with caution.

So why are auction results down in 2016?

Another particular characteristic of the art market is that the assets being sold are also consumer goods and people have positive emotional returns from buying a painting, other than pure investment. This blurred line between collecting and investing makes the demand, people buying art, behave differently from other assets, making comparisons harder. For instance, people who buy art usually do it because they have extra money to spend, diminishing liquidity concerns. If a bubble were to burst, collectors wouldn’t need to sell their works right away to pay for the rent, causing a “herd effect” on the market. The “herd effect” is a very common consequence of bubble bursts, for example when a rumour goes round about a bank going broke. In this case, everyone with money in the bank will run to withdraw their savings, leaving the situation even worse. There is very little risk of this happening in the art market.

This doesn’t mean that the market can’t be overvalued. Having prices more expensive than what assets are actually worth doesn’t mean that there is a bubble, just that prices will eventually converge to the established level or lower. A bubble burst would plummet all the prices in a short period of time. The term “correction” is used to describe a market that, for a number of reasons, has grown faster than it should and is now softening to a more plausible level of prices. This is what is happening in the art market.

The key to understanding the growth in sales in the last decades, besides the growth of the 0.001% richest worldwide, is the entrance of new Chinese players in the market. The Citi report stated that 33% of all revenue growth came from new Chinese collectors, supported by a government expansionary monetary policy. Now that the Chinese miracle is almost over and the economy doesn’t show the same signs of vertiginous growth as in previous years, the art market is suffering. We could compare this situation to a town where a big industry decides to set up a plant. Because many new employees move in and have to find apartments in the city, housing prices rise to a new level. As the factory keeps growing, prices go even higher. When economic factors cause the company to stop hiring new people and even dismiss some staff, apartment prices will stay at the old level or even decrease in value a bit. This is a correction, not a bubble. But if the new employees buy new apartments far more expensive than what they can afford, artificially inflating the prices, and for which they now lack the money to pay, this is a bubble. So unless the Chinese buyers leveraged themselves to buy the artworks, there’s no reason to believe they have caused a bubble and could potentially burst it.

The overall economic environment is also detrimental to other new collectors. Oil prices are extremely low and are not expected to rise again in some years, hurting EMEA and Russian collectors. The latter have seen their currency dramatically devaluated and the signs of recession on the Russian art market are clear. US and European economies have not yet recovered and this art market correction to lower price levels, visible in 2016, actually seems to be the current trend, but is very different from a bubble burst.