An investment flash: the rate of return for photographs.

From today, painting is dead. Paul Delaroche, 1839, upon seeing a daguerreotype for the first time

I. Introduction

Although the general public has not completely recognized photography as an investment alternative, art collectors are snapping up the work of photographic masters at record prices. In October 1993, a vintage photograph by Alfred Stieglitz, Georgia O'Keeffe: A Portrait -- Hands with Thimble, sold for $398,500 at Christie's International auction house.(1) This bettered the previous record price of $193,895 set in May 1993 by a Man Ray photograph. Even if these prices do not approach the millions of dollars paid in recent years for work of impressionistic painters, these record prices have attracted investors' attention.

The objective of this study is to examine the investment value of photographic prints. The market for collectible photographs boomed in the 1980s as investors became aware of the value of photographs as an investment, especially as an alternative to the high prices of paintings and sculpture. Although various studies have investigated the investment value of collectibles such as paintings [1; 2; 9; 18; 26], prints [20; 22], coins [6], Stradivarius violins [24], wine [16], and antique furniture [12], there have been no detailed studies of the performance of photographs as financial assets.

The research completed generally concludes that paintings and prints are a poor long-term investment. Baumol [2] calculated the average annual real rate of return for painting at .55 percent, concluding that art was a "floating crap game," and a poor investment alternative in general. Frey and Pommerehne [9] extended Baumol's study and found an average real rate of return of 1.5 percent. Stein [26] calculated the geometric average real rate of return for paintings sold in the United States and United Kingdom between 1946 and 1968 to be about 8 percent. Pesando [22] estimated a mean real return of 1.51 percent for prints between 1977 and 1992. Mok, et al. [18] find a high yield of 52.9 percent for Chinese paintings from 1980 to 1990, although they conclude that investing in stock market portfolios offer better alternatives.

II. The Market for Photography

Despite Delaroche's ominous and inaccurate pronouncement on the demise of painting, photography was not immediately accepted as a true art form. Indeed, the debate over whether photography was art continued for many years after Daguerre first made public the photographic process in 1839. It was not until 1859 that photographs first won the right to be exhibited as "art" next to paintings [8, 1]. Peter Henry Emerson in the 1880s and Alfred Stieglitz in the first half of the 20th century persuasively argued that photography was a legitimate art form. The acceptance of photographs as art may have lagged because the technology and chemistry used to create a photograph make the process seem impersonal and mechanical. With modern technology providing numerous ways (e.g., digital imaging) to produce visual images, a work of art is now more accepted for the significance of the final image, rather than the process (e.g., computers) used to achieve the end result. Today, photographs appear in nearly all of the major museums and galleries, and the question as to whether or not photography is a legitimate art form has clearly been answered in the affirmative.

Although the legitimacy of photography as art has been resolved for some time, it was not until 1975 when major auction houses began to regularly schedule fall and spring photography sales. The development of the photographic art market in the late 1970s was coincident with the growth in the number of photographic exhibitions at influential museums. In the 1979-80 auction year, of the total of $434 million in art sales at the twelve largest auction firms in the United States, slightly over 1 percent of total sales were for photographs, a 100% increase over the previous year [8, 7]. The market for photography boomed in the 1980s, partly as a result of the inflated prices that were being paid for paintings and sculpture.(2) Prices are generally more affordable for photographs, and some works have shown remarkable price gains over recent years. For example, a set of The Native American Indian by Edward Curtis sold for $650 in 1953 was purchased for $662,500 in 1993 [10].

Although similar to the market for other art objects such as painting and sculpture, photography has characteristics that make it different. Theoretically, unlimited numbers of photographs could be created from a single negative, making the price for any single photograph much less than the price paid for a one-of-a-kind painting.(3) The possibility of unlimited use of a negative to produce numerous prints is of major concern to collectors, since the importance of scarcity is recognized by potential buyers. A Richard Avedon photograph, "Dovima," sold for $19,800 in 1991, partly because the negative had been destroyed [19].

In addition to the photographer, the printer of a negative can affect the value of the photograph. For example, an Edward Weston negative printed by the photographer is more valuable than the same negative printed by Brett Weston, Edward's son. Less expensive still are posthumous Edward Weston prints done by Cole Weston, another son [4, 16]. In general a photograph is most valuable if the photographer used the original negative and made the print at the time the negative was taken. Other factors that determine the value of a photograph include print quality, importance of the photographer and of a particular photograph within the photographer's work, beauty of composition, physical condition, authenticity, rarity, vintage, history, signature, stamp, date, notes of photographer, presence in prominent collections and exhibitions, and prices achieved at auctions [27].

An index to track the market for photographs was created by the staff of The Photographic Arts Center (TPAC) [21].(4) The index, which is computed in a manner similar to the Dow Jones Industrial Average, follows the sale of a selection of photographs representative of the market, based on a photograph's historical importance (e.g., daguerreotypes) and aesthetic value, as determined by the staff of TPAC. For the period between 1975 and the fall of 1995, the index showed an increase of 1680%, while the Dow increased by 587% over this period. The index provides a broad measure of the investment potential of photographs.

III. Data Set and Methodology

For this study, data on photographic prices and characteristics are taken from the 1980 through 1992 editions of the Photographic Art Market. These publications list all photographs sold at the major auction houses (i.e., Sotheby's, Christies, Swann, and Butterfields) in the United States between 1980 and 1992, providing information on location of sale, date of negative, date of print, size, and price. Of the approximately 37,400 sales listed for sale during this period, almost 72 percent were sold, with the remaining works "bought in".

Following the rationale of TPAC, a subset of 25 photographers or groups of photographers (e.g., daguerreotypes of various photographers) is collected from the sales of all photographs to form the data set. The categories and photographers in each category are listed in Table I. The six broad categories selected by the staff of TPAC as representative of the photographic art market, are two categories of European photography--the nineteenth (EUR19) and the twentieth century (EUR20), and four groupings of American photography--the nineteenth century and the West (AM19), early twentieth century and photo-secession (AMEA20), the 1920s and 30s (AM2030), and the 1940s to the present (AMPR).

Table I. Categories and Photographers

Category                          Symbol   Photographers

European, 19th century            EUR19    Emerson, Cameron,
                                           MacPherson, Bisson
                                           Freres, Baldus,
                                           Hill and Adamson
American, 19th century and West   AM19     Muybridge, Curtis,
                                           Watkins
American, early 20th century      AMEA20   Atget, Stieglitz, Strand
  and Photo-secession
European, 20th century            EUR20    Kertesz, Brassai, Man
                                           Ray, Cartier-Bresson
American, 1920s and 1930s         AM2030   Abbott, Weston, Lange
American, 1940s to present        AMPR     Adams, "Weegee," Penn,
                                           Karsh, Greene, Dater

The repeat sales technique, a method used in similar studies investigating the investment value of collectibles [2; 9; 18], is employed here. Using this method, an observation is recorded when the exact photograph is sold on two different occasions.(5) To estimate the holding period return (HPR) for a particular photograph the following formula is used:

HPR = ([P.sub.2] x .9)/[P.sub.1]

where P1 and P2 are the purchase prices at the first and second auctions, deflated by the Consumer Price Index using the 1982-84 base period. [P.sub.2], the actual hammer price, is multiplied by .9 to adjust for the buyer's premium of 10 percent, to arrive at the amount received by the seller.(6) The annual holding period yield (HPY) is then calculated and adjusted for risk by dividing by the standard deviation of the HPY [23].

IV. Results and Analysis

Using only repeat sales leads to a data set of 1192 observations. Table II lists the average real prices, real rates of return, standard deviation, and adjusted real rates of return, disaggregated into the six photographic categories defined above. The average price of a photograph is $1969, with the average price per season rising from $1802 in the 1980-81 season to $3301 in the 1991-92 season. Prices range from $41 to $26,080, while the average price for a photograph in any period is under $3000. The average length that a photograph is held is 1.9 years.(7) The average annualized rate of return is .302, or approximately 30%.(8) Slightly more than 1/2 of the returns (623) are negative.

Table II. Annualized Real Rate of Return and Risk-Adjusted Rate of
Return for Photographs, 1980-1992

         Average   Rate of   Standard    Adjusted Rate
Period    Price    Return    Deviation    of Return

EUR19    $ 1319    .120       1.469        .082
AM19       2373    .402       2.024        .199
AMEA20     2166    .236        .976        .242
EUR20      1684    .371       3.782        .098
AM2030     1514    .344       2.282        .151
AMPR       2561    .209       3.077        .068
AVE        1969    .302       2.980        .101

           Number of
Period   Observations

EUR19       48
AM19        87
AMEA20      48
EUR20      371
AM2030     271
AMPR       367
AVE       1192

Examining the six categories is of interest, although results should be viewed cautiously since individual photographers within categories can vary significantly from the average. There are fewer nineteenth and early twentieth century prints available for sale, so that these categories--EUR19, AM19, and AMEA20--have the fewest number of observations. The yields vary across the six periods, with the highest return of .402 for AM19. Of the two photographers representative of this period, Muybridge photographs offer high returns of .688, while Curtis is below average at .209.(9) The lowest return of .120 is for EUR19, which also has the lowest average price of $1319. This may be partially explained by the fact that EUR19 and AM19 have more observations than any other period sold in the early years (1980 to 1983) and the later years (1989 to 1992), respectively of this data set. The highest average price, $2561, is realized by photographs of AMPR, although this period has lower than average returns. Three of the photographers in this period--Penn, Adams, and Greene--are among the lowest individual returns.

As indicated by the high standard deviations, investing in photographs can be risky, especially for certain periods. For example, although EUR20 offers the second highest return at .371, it has the highest standard deviation. The category for recent American photographers, AMPR, has a lower than average return and a high standard deviation of 3.077. The returns can be adjusted for risk by dividing by the standard deviation to get the risk per unit of return. Doing so, gives an average return for the total data set of .101. The relative position of photographs in the period AMEA20 improves significantly. This period's photographs, produced by Atget, Stieglitz, and Strand, have the highest adjusted return of .242, and the smallest standard deviation at .976. A relatively good return is also earned by AM19 which is the highest unadjusted, and the second highest adjusted yield. The categories EUR19 and AMPR, have low rates of return, both adjusted and unadjusted. The lowest adjusted return is for AMPR at .068, which is next to the lowest when unadjusted. Work from EUR19 is the second lowest adjusted, and the lowest unadjusted.

Since some photographers are heavily represented in certain periods, it is useful to consider some individual returns. For example, in the latter three categories, EUR20, AM2030, and AMPR, a single photographer in each category (Cartier-Bresson, Abbott, and Penn, respectively) has a significantly greater number of observations than any other. As would be expected, there is significant variation among the rates of return for individual photographers, listed in Table III by the six periods. Highest average returns for individual photographers are earned by Kertesz (2.09), Weston (1.416), Dater (.937), and Muybridge (.688). The average price of a Dater photograph of $757 was among the lowest average prices for a photographer.(10) Lowest returns for individual photographers were for Baldus (-.628), Adams (-.09), Brassai (-.09), Penn (.03), and Emerson (.05). From Table III it is clear that within each period, photographers differ considerably above or below the average. It is wise, therefore, to consider individual photographers, rather than a particular period, when making an investment decision.

Table III. Annualized Real Rate of Return for Individual
Photographers
                                                    Number of
                   Mean Price    Average Return    Observations

Emerson              1353             .051              16
Cameron              1637             .150              21
MacPherson            288             .320               2
Bisson Freres         540             .358               2
Baldus                561            -.628               1
Hill and Adamson      941             .178               6
Muybridge            2203             .688              35
Curtis               2488             .209              52
Watkins                 0                0               0
Atget                2479             .372              23
Stieglitz            3241             .093              11
Strand                867             .234              14
Kertesz              1949            2.092              28
Brassai              1561            -.050              16
Man Ray              5794             .136              10
Cartier-Bresson      1563             .249             317
Abbott               1266             .312             262
Weston               9802            1.416               8
Lange                 206             .251               1
Adams                5165            -.090              55
Weegee               1114             .172              64
Penn                 3603             .030             136
Karsh                1057             .328              36
Greene                701             .081              19
Dater                 757             .937              57

The highest average prices for an individual photographer were for Weston ($9802), Man Ray ($5794), and Adams ($5165). The low returns for Ansel Adams are curious, since he is probably the photographer most recognized by the general public. Prices for his work appear to have peaked early, so that upon resale the photographs may offer a disappointing return. It may be that since Adams is the name most recognized by buyers, the uninformed buyer may pay prices in excess of the true market price.

Although this study finds that the average annualized rates of return for photographs are generally higher than rates of return for other works of art, returns may not be as high in the future. Since the market for photographs "took off" in the 1980s, one would expect returns for this period to be high. Examining shorter time frames within this period offers some evidence that returns for photographs may be levelling off. For instance, the average annualized return is 33.3 percent between 1984 and 1992, 26.1 percent between 1986 and 1992, 23.8 percent between 1988 and 1992, and 21.4 percent between 1990 and 1992. Although the average annualized returns do not decline for every year, or every period, a declining trend is evident.

Also, the high returns for art in general through much of the 1980s may be partially responsible for the high returns for photographs. The Daily Telegraph Art 100 Index (DT 100), a yearly index for sales of paintings at auction houses, can be used to determine if returns to paintings and photographs are growing closer [7]. For the period from 1980 to 1992, the Index increased by an average of 17.7 percent annually. Comparing paintings and photographs for these periods shows that the returns for photographs decreased by a slower rate than paintings. Prices on paintings increased steadily through the 1980s until the correction of the early 1990s. The DT 100 Index increased yearly by an average of 20.82 percent from 1984 to 1992 and 13.48 percent from 1986 to 1992. However, the DT 100 decreased yearly by an average of 2.87 percent from 1988 to 1992 and 16.76 percent from 1990 to 1992. The decline in the Index by over 40 percent from 1990 to 1991, and a 16 percent decline for the following year show the collapse of sales of painting in the early 1990s. Although rates of returns on photographs declined in the late 1980s and early 1990s, the decline was much more modest than for paintings.

Finally, some of the high returns to photographs may be due to the short time period of the data set.(11) Buelens and Ginsburgh [5] note, that over the short run, some schools of painting might offer much higher rates of return than other schools.

Some general cautions about art markets also apply to photographs. Auction sales only show price fluctuations of art in demand, so that an upward bias may exist [11,13714. As an investment vehicle, photographs are subject to some of the same risks and costs as paintings. A work of art is subject to problems of deterioration (e.g., fading or oxidation) or destruction. Art is not as liquid as some other assets, thus increasing the financial risk [17, 54]. These costs along with insurance costs must be deducted from the rate of return to make an accurate comparison between photographs and other investment vehicles such as stocks and bonds. However, the high yields for photographs compare favorably to some traditional assets. For example, between 1975-90, the annual average rate of return for U.S. Government securities was 2.54 percent, while stocks averaged 8.14 percent [14, 165]. Although these returns are less than returns for photographs, the risk is less also.

If markets are efficient, the return on photographs, net of costs, uncertainties, and consumption value, should approximate the return to other investment alternatives. Since the photographic art market is fairly recent, and this is the first study of this type, investors should view the results of this study with caution. One should study the market experts (4; 8; 21; 28) for additional information.

V. Concluding Comments

This paper examines the performance of photographs, an important investment vehicle, but one about which very little is known. The photography market is still young and is subject to fluctuation, with particular works likely to exhibit major price swings as the market adjusts. However, a substantial market for photography has developed, and good investment opportunities exist. This study finds that the market for photographs offered high returns over the period from 1980 to 1992. However, since the art market in general saw escalating prices over most of this period, these results may be biased upwards.

Art investment is often viewed as an area where the investor is likely to earn spectacular gains. This popular view is often voiced when particular works of art, such as Van Gogh's Irisis, are sold for millions of dollars. Studies by economists [1; 2; 9; 26] indicate instead that lower average returns are earned on art compared to other alternative assets. This is due in part to the fact that art, unlike other kinds of traditional investments, provides "consumption value," in addition to asset value [13, 153]. That is, the art investor enjoys the aesthetic and prestige value of the art, while at the same time possibly earning a positive return on the investment. Given the risks and transaction costs involved when investing in art, investors are ill advised to use art solely as an investment vehicle. However, the art market may become an increasingly important part of a varied investment portfolio. In a time series analysis of the art market, Goetzman [10], finds that in the last 50 years there has been unprecedented growth in art prices.

Squires [25] notes that collectors of photographs have become knowledgeable only since 1990. Therefore, some of the prices paid up to this date may not accurately reflect the true value of the photograph. Although some photographers have provided very good rates of returns, based on the current study, trends can change abruptly and often, so that past performance may not be a good predictor of future performance. However, with the majority of collectible photography selling for under $10,000, photography presents an opportunity, especially for the small investor, to compile an affordable but important art collection. (*) I would like to thank the Faculty Research committee from Francis Marion university for financial support. Useful comments were provided by Robert Persky, James Rinehart, James McFadyen, and an anonymous reviewer.

(1.) A vintage print is one that is developed by the photographer, around the same time as the date of the negative. This is in contrast to negatives that are often developed and printed later, sometimes by someone other than the original photographer. (2.) The high art prices of the 1980s, fueled by pressure from Japanese speculators, collapsed in the early 1990s, falling by 50 to 75 percent, as Japanese investors fled. Some signs indicate that by 1995 the art market may be recovering [3].

(3.) Prior to 1970, photographers seldom produced more than ten prints from a single negative [8, 25]. Recently, some photographers have begun to practice editioning--making only a limited number of prints from a negative, similar to limited editions of etchings.

(4.) The index is calculated by using a group of 25 photographers or groups of photographers that are representative of the market. If a particular image is not sold in a given auction year, a comparable image is chosen.

(5.) Although a print from a negative could be printed on more than one date, only photographs that were printed on the same date are included. Prints from the same negative could be slightly different, e.g., lighter or darker, thus leading to a difference in value for reasons other than market conditions.

(6.) For the 1980-81 auction year, since the price listed did not include the 10% buyer's premium, [P.sub.1] was multiplied by 1.1 to get the actual hammer price

(7.) The average holding time for the sample is relatively short. Mok, et al. [18] suggest that if a work is sold after 2 years, it's annual return may be less than if it sold after 4 years. However, Buelens and Ginsberg [5,1354] show that rapid resale does not introduce bias.

(8.) If the photo-secession magazine Camera Work published by Stieglitz is included, the data set expands to 1309. Including Camera Work leads to slightly lower returns. For example, the average annualized rate of return is .278 rather than .302.

(9.) No observations are recorded for Watkins, a member of this category.

(10.) The highest rate of return for an individual observation is 56.57 for Dater's Imogen and Twinka at Yosemite, which sold in October 1992 for $941.16 and in November of 1992 for $1568.60.

(11.) A short time period avoids some problems that studies of art over longer periods have. Holub, Hutter, and Tappeiner [15] criticize studies of art [2; 9] that use long periods of data.

References

[1.] Anderson, Robert C., "Paintings as an Investment." Economic Inquiry, March 1974, 13-26.

[2.] Baumol, William J., "Unnatural Value: or Art Investment as a Floating Crap Game." American Economic Review, Papers and Proceedings, May 1986, 10-14.

[3.] Berman, Ann E., "Back to the Future: '80s Prices for Art." Wall Street Journal, May 16, 1995, A16.

[4.] Blodgett, Richard. Photographs: A Collector's Guide. New York: Ballantine Books, 1979.

[5.] Buelens, Nathalie and Victor Ginsburgh, "Revisiting Baumol's `Art as Floating Crap Game.'" European Economic Review, 37, 1993, 1351-63.

[6.] Dickie, Mark, Charles D. DeLorme, Jr. and Jeffrey M. Humphreys, "Price Determination for Collectibles: The Case of Coins." Southern Economic Journal, July 1994, 40-51.

[7.] Duthy, Robin. The Daily Telegraph Art 100 Index. London: Art Market Research, 1995.

[8.] Falk, Peter. The Photographic Art Market. New York: Falk-Leeds International, Inc., 1981.

[9.] Frey, Bruno S. and Werner W. Pommerehne, "Art Investment: An Empirical Inquiry." Southern Economic Journal, October 1989, 584-92.

[10.] Gates-Warren, Beth and Denise Bethel. Twenty Years of Photographs at Sotheby's. New York: Sotheby's, 1995.

[11.] Goetzman, William N., "Accounting For Taste: Art and the Financial Markets Over Three Centuries." American Economic Review, December 1993, 1370-76.

[12.] Graeser, Paul, "Rate of Return in American Antique Furniture." Southern Economic Journal, April 1993, 817-21.

[13.] Grampp, William D. Pricing the Priceless: Art, Artists, and Economics. New York: Basic Books Inc., 1989.

[14.] Heilbrun, James and Charles M. Gray. The Economics of Art and Culture. New York: Cambridge University press, 1993.

[15.] Holub, H. W., M. Hutter, and G. Tappeiner, "Light and Shadow in Art Price Computation." Journal of Cultural Economics, 17, 1993, 49-69.

[16.] Jaeger, Elizabeth. "To Save or Savor: The Rate of Return to Storing Wine." Journal of Political Economy, No. 3, 1981, 584-92.

[17.] Keen, Geraldine. Money and Art: A Study Based on the Times-Sotheby Index. New York: G.P. Putnum's Sons, 1971.

[18.] Mok, Henry, Vivian Ko, Salina Woo, and Katherina Kwok, "Modern Chinese Paintings: An Investment Alternative?" Southern Economic Journal, April 1993, 808-14.

[19.] Peers, Alexandra, "Picture This: Buyer's Market Develops for Photo Prints but It's Buyer Beware." Wall Street Journal, October 4, 1991, C1.

[20.] Penn, Robert E., "The Economics of the Market in Modern Prints." The Journal of Portfolio Management, Fall 1980, 25-32.

[21.] Persky, Robert, editor. The Photographic Art Market Auction Prices, 1980-1992. The Photographic Arts Center: New York.

[22.] Pesando, James E., "Art as Investment: the Market for Modern Prints." American Economic Review, December 1993, 1075-89.

[23.] Reilly, Frank K. Investment Analysis and Portfolio Management, fourth edition. Hillsdale, Illinois: The Dryden Press, 1994.

[24.] Ross, Myron H. and Scott Zandervan, "Capital Gains and the Rate of Return on a Stradivarius." Economic Inquiry, July 1989, 529-40.

[25.] Squires, Carol, "Elements of Collecting." American Photo, March/April 1995, 53-57, 84-85.

[26.] Stein, John P., "The Monetary Appreciation of Paintings." Journal of Political Economy, August 1977, 1021-35.

[27.] Studer, Margaret, "Art Investing is Enlivened by Photography." Wall Street Journal, March 3, 1993, C1.

[28.] Witkin, Lee D. and Barbara London. Photographs: A Collectors Guide. Boston: New York Graphic Society, 1979.