Spring Oriental Rug Auctions; How Relevant Are Estimates?
by George O'Bannon
Skinner's, Christie's, and Sotheby's had Oriental rug sales in April. It seemed that
rug dealers were out in force at all of them, even many of the Europeans. But the mood at
the sales continues to be very business-like with a lack of enthusiasm. There was,
however, one auction of rugs that did generate a lot of enthusiasm, the Nureyev sale in
January. Perhaps what the rug world needs is a world-famous personality.
Something that many auction attendees don't fully appreciate is the reliability of
estimates placed on auction lots. Most people assume there is a high degree of
reliability. Auction catalogs explain the intent to potential buyers: "The estimates
are guides for prospective bidders and, where possible, reflect prices that similar
objects have sold for in the past. The estimates are determined several months before a
sale and are therefore subject to revision to reflect current market conditions. Estimates
should not be relied upon as a representation or prediction of actual selling prices"
(Sotheby's). Christie's statement is much the same, and Skinner's has no estimate
definition in its catalog. Basically, estimates are guides; don't count on them. You have
to decide what you are willing to pay.
Many prices paid for pieces pointed up the vagaries of estimates in the recent sales.
Some were realistic, some very unrealistic, some set deliberately low to attract bidders.
Who is at an auction to buy or not buy affects the estimates' reliability, and some
estimates couldn't be counted on at all because of the celebrity nature of the goods.
Let's begin with the celebrity one first, the Nureyev sale.
M.A.D. readers were informed of the enormously successful and near hysterical
auction on January 12-13 of Rudolf Nureyev's estate in the April issue. If proof were
needed that celebrity can make people spend excessively for personal items at auction,
this one proved it in spades. The catalog, printed in hardcover, sold out before the
auction and is likely to become a collector's item.
The lots that brought the most ridiculous prices were his personal items and costumes
from his ballets. But the things he collected, including paintings, sculpture, furniture,
prints, maps, textiles, rugs, and photographs, also surpassed their estimates. One had the
sense that Nureyev was not just an accumulator. It seemed he acquired a variety of arts
from the European tradition of keeping one's wealth in different kinds of assets against
the day when hard times return, when things had more value than money. With few
exceptions, they were not bought with a focused collector's eye or knowledge.
This was particularly true of the 64 lots of kilims. Most were in excellent condition
or had been restored to excellent condition. But they ranged from early 19th-century
pieces with great color and design to more recent ones with obvious synthetic dyes and
uninspired weaving. Two thirds of these lots sold below or within estimates, which
occurred with no other group of items in the auction. Since many of the lots consisted of
several rugs, the hammer prices in most cases were actually bargain-basement.
The reason for this would seem to be that the kilims did not have the close personal
association with Nureyev that other items had. For instance, the Kashmir shawls, which he
is known to have worn from photo documentation, had his "scent" on them. The
kilims did not. All of the shawls sold above estimates and, in many instances, more than
doubled the high estimates.
But this was a sale for balletomanes and fans. Few rug dealers showed up, in large part
because the kilim market exists as a separate and specialized entity within Oriental rug
sales. There are two markets for these weavings, decorative and collector. For the
decorative market, kilims are readily available at wholesale; the collector market is
highly selective and in recent years thin and erratic. There was no reason for dealers to
consider buying for resale in a situation where hysterical fans might pay anything, nor
were the pieces of the quality sophisticated kilim collectors are buying, in these
post-"the Mother Goddess is dead" days.
The first lot offered, 335, was an unusual size for an Aksaray kilim and had a white
cotton field, also unusual, which gave the bold motifs great clarity. It sold for $1150
(all prices include buyer's premium) against an $800/1200 estimate&emdash;a good buy
for personal use by a prominent textile dealer. Lot 337, one of the more exceptional
pieces, was a rare format for this type of stacked prayer niche design (est. $2000/3000).
It sold to an absentee bidder for $4025, a reasonable retail level.
Lot 372 showed why dealers have problems selling many Turkish kilims. The 14 feet 1
inches x 6 feet 3 inches size is too long and narrow for most people to use. Although a
good late 19th-century piece with striking, bold patterning, it brought only $1840 (est.
$2500/3500). Lot 381 was one of the best buys in the sale at $920. This piece was
double-woven, a rare technique for this type of kilim. Retail on such a piece would be
around $3000, if you can find one. It had a very low estimate at $800/1200 and also
included a second kilim.
Lots 384 and 390 were Thracian kilims in near-square sizes, 12 feet 10 inches x 11 feet
and 12 feet 2 inches x 10 feet 4 inches, of a desirable and decorative type, and lot 390
was unusual in having rows of women around the edge of the field. The estimates of
$2000/3000 each were ridiculously low&emdash;$6000/8000 would have been more in
line&emdash;and they sold for $8625 and $12,650.
Lot 396 was one of the oldest kilims in the sale, probably from the first half of the
19th century and an unusual format. The $2760 price (est. $1200/1800) was a wholesale one
for this rug.
For kilim buyers, the Nureyev sale was one not to have missed. If there were any buys
to be had in this sale, they were it.
Of the major auctions, Skinner's April 8th sale was the most successful on a percentage
of lots sold basis (88. It was well attended (there was a full house), and people seemed
to be there to buy. A bank of seven telephones (the most Skinner's has had) had been
installed for phone bidders. It seems the Boston sales attract many dealers for the
preview who then bid on the phone. Skinner's has also found that the Boston galleries
attract a larger number of private buyers than in Bolton.
There were no spectacular lots, but there were the usual types of good, small pieces
for collectors. One new collector was able to pick up three very nice pieces&emdash;a
Baluch bagface (lot 83) for $575 (est. $300/500), a Veramin bagface (lot 122) for $747.50,
and a Yomud torba (lot 150) for $690. All of these were mainstream collector items that
have been sought after for many years now.
A southwestern dealer, new on the auction scene, bought eight pieces for inventory,
covering a wide range of types. There were many new dealer faces, and one wonders if they
are attending auctions because fewer and fewer goods are coming their way through normal
Lot 24 was a very good example of a dragon sileh. It showed considerable wear but was
unusual in being woven in one piece instead of the more normal two panels. The drawing of
the dragons was of a type considered early, and it sold to the phone for $4025 (est.
$2000/2500), a great buy. With $500 to $700 of restoration, it will be a great piece for
use or resale. In contrast, a two-panel dragon sileh at Sotheby's (lot 67), in better
condition but showing obvious synthetic dyes, sold for $7475.
Skinner's has always had good Caucasian pieces, and four of the best were pictured in
color in the catalog. Although still popular with buyers, prices have come down
significantly in recent years. Lot 88, a Pinwheel Kazak, sold at what must have been its
reserve, $8912 (est. $10,000/ 12,000). Although showing some wear and thin areas, this rug
would have brought $15,000 five years ago.
The best Kazak in the sale, lot 90, had a full pile and beautiful wool and patina. It
reached $6500 (est. $8000/10,000) and was bought in, but it sold post-sale at $8050.
Another Kazak, lot 94, was not as interesting because of the main border pattern and the
whiteness of the field, but it was in excellent condition and sold for $8050. Lot 96, a
Talish rug, had great charm because of a couple holding hands and other patterns on the
open, abrashed blue field. The crab design on an ivory ground for the main border was very
dramatic. It sold for $5750 (est. $3000/4000) and was a great buy.
Turkoman and Baluch lots were one of the weak segments of the sale, primarily due to
the quality of the goods. But lot 98, a Tekke kapunnuk, caught the interest of several
bidders as it went well above expectations at $12,650 (est. $6000/8000). A local collector
and phone bidder went to war on it, and the phone won. Although in very good condition and
with good drawing, one of the arms was a few inches longer than the other, which was a
One can't help but comment on lot 159, the 17 feet x 11 feet 4 inches Tabriz carpet
with an Ardebil design. The original Ardebil, purchased in 1893 by the Victoria &
Albert Museum in London, is extolled in many rug books as the most famous and finest
Persian rugs ever made. There are many who would debate that judgment, but it is a
beautiful rug and design. It has been copied by weavers in most of the major rug weaving
regions throughout the 20th century. This particular copy was one of the most atrocious
ever seen. The scale of the design and patterns was too small, the white ground of the
field was too white, it had a green in particular that was the gaudiest, and other
unmentionable features. It was the type of rug that gives "the finest Persian
carpets" a bad name. Mercifully, it was bought in at $10,000 (est. $15,000/20,000).
Sotheby's had a similar piece, smaller in size, that sold for $26,450.
In contrast to such pieces, lot 91 was a beautiful antique Sarouk from the end of the
19th century. The sparsely patterned rose field balanced the elegant blue medallion and
ivory spandrels perfectly. The $13,800 price was a very good buy, and it would be a joy to
live with, not a copy of something else, and expressive of the Persian village idiom.
Frequently, the auctions will offer lots of books. In this sale there were four lots of
Hali magazines, which were a good example of how the prices for them can vary.
Lot 28 included 33 issues from the first nine volumes of Hali, including Vol. 1,
No. 1, which can bring $600 or more at auction if offered singly, and sold for $2990 or
around $90 per issue. Lot 32 also consisted of 33 issues, but from volume three through
volume nine, and sold for $1092.50 (est. $1500/2000) or close to $33 per issue. Lot 30
included only the four issues of the rare volume one; estimated accurately at $1000/1200,
it sold for a bargain $575 or $143.75 per issue. Lot 31, which included only volume two,
sold for $460 (est. $700/900) or $115 per issue.
Although different people bought all four lots, if one had bought lots 30, 31, and 32,
one would have had a better deal than lot 28 and ended up with more issues, including
volumes one and two. C'est la vie, one never knows how these things will turn out.
Back issues may be bought new from Hali at $120 for volume five, $110 for volume
six, $100 for volume seven, $80 for volume eight, and $50 for volume ten for a retail
Skinner's next major Oriental rug sale will be held on Saturday, September 16. A summer
sale was held on July7.
Of all the rugs offered in the spring trio of auctions,
Christie's April 10th  sale had the most important carpet. Lot 100, a 12 feet 8
inches x 13 feet 6 inches Mughal millefleur "star lattice" carpet, had been sold
at Rippon-Boswell in Germany in November 1989 for $726,000 and set the record for an
Oriental carpet at auction. The carpet was from the Vanderbilt estate and had hung in the
Moorish smoking room of the Vanderbilt mansion in New York City. Christie's published a
photo from the New-York Historical Society archives showing the carpet on the wall.
For a late 17th-/early 18th-century Moghul carpet, it was in
excellent condition. Few such rugs remain in private hands, and this was a rare
opportunity to examine and handle a rug of this type and quality. It attracted the
attention of the European dealers in particular, who have the clients with the taste and
money for this type of rug. There were questions about why it was being returned to the
market so soon, and the $700,000/900,000 estimate seemed to show that the owner was
willing to take a loss. The general feeling was that it was going to be hard to sell in
When the bidding opened, there was no interest in the room as
the auctioneer kept increasing the amount to what appeared to be the reserve, $650,000. At
that point two phone bidders came into play. The bidding increased slowly with occasional
hesitation by the phones about going higher. It inched up, and the carpet sold for
$992,500. There was mild applause, one almost sensed relief, as the gavel fell. The rug
reportedly had been bought by a European at Rippon-Boswell, and it is returning to Europe.
Lot 102, an 11 feet 11 inches x 8 feet Sarouk Ferahan, was an interesting example of
how classical carpets such as the Mughal carpet were obvious inspirations for later
weavers. Although certainly less dramatic, this carpet had the same star lattice design
filled with similar floral forms. It sold for $16,100. Here, one could say not a copy but
"in the spirit of."
Although most types of rugs have names that convey relative certainty about where or by
whom they were made, there are some types still poorly identified. Kurdish rugs are
probably the most notable of these. Some weavings are attributed to specific Kurdish
weavers, but most Kurdish rugs are called simply Persian and Turkish Kurd. A relative
recent specific Kurdish attribution is Kuchan Kurd, meaning the group of Kurds near that
town in northeastern Iran. Lot 108 was attributed to them, apparently on the basis of
color and simplicity of design, but the long pile was longer than normal for the group; it
was more like the yatak or julkhirs sleeping rugs from Turkey and Central Asia. Central
Asian julkhirs rugs are one of those still poorly understood groups, and this lot was more
likely to be an Uzbek julkhirs than a Kuchan Kurd. The weaving was passed at $2400 (est.
$3000/ 4000). Too strange to be bought as a Kuchan Kurd and unknown as an Uzbek julkhirs,
it may have to wait a few more years to be appreciated.
This sale grossed $2,156,150 with 56of the lots selling&emdash;not a very good
result, particularly if the Mughal carpet amount is deducted. The main weakness seemed to
be in a large number of unsold decorative carpets. Christie's next sale is scheduled for
This sale on April 13 featured 47 lots from a private collection, which frequently
makes for some very exciting bidding and prices. Lots 1 through 47 were from Mr. Benjamin
Sofer of Israel and consisted primarily of Central Asian suzanis, embroidered bed covers
and hangings, and rugs and kilims. It was reported that Mr. Sofer, well known in Israel
for his collection, had recently suffered a theft with the best pieces being stolen. For
that reason he had decided to sell out completely. The estimates on most of these pieces
were lower than one would normally have expected.
Lot 1 was a Tekke Turkmen embroidered wedding trapping with a beautifully rendered
pattern of five flowering plants in the field and horses and wedding participants in the
top border, a rare feature on these pieces. It has been years since a similar example
appeared in the auction market, and the $5000/7000 estimate seemed reasonable. There
turned out to be many bidders interested, and after rapid-fire increases, it sold for
$11,500&emdash;a very healthy start.
Lot 9 was a suzani, probably from Bokhara, in a typical bed cover size of 7 feet 7
inches x 3 feet 2 inches. This piece had a common lattice with flowers on the field and
floral meander for the main border. But the large palmettes of the border and lattice were
delicately drawn, and it sold to the phone for $5175. Lot 13, a Tashkent suzani with large
red discs for patterning, termed palak, was unusual for the type because of several
squares without palak patterns and brought $6900 from a left bid. Lot 36, related in style
to lot 9 but unusual in having scalloped niches at the ends of the field, sold for $5462
to an American collector and authority on these weavings.
It was said that these were the lesser of Mr. Sofer's pieces because of the theft. One
can see that in comparing lot 38, a 2 feet 2 inches x 2 feet 2 inches suzani called a
bokche, to another Sofer piece published in color in Yanai's book, fig. 18 (see reading
suggestions below). Because of their size, these pieces are almost always charming, but
the embroidery was a bit too dense in this piece. As is often the case, small pieces cost
more per square feet than larger ones, and it sold for $5175.
The same American collector bought lot 43, a small (5 feet 3 inches x 3 feet 8 inches)
suzani with a free, unrestrained flowering plant in the field and undulating palmettes and
leaves in the border. It sold for $8625 (est. $4000/6000). The major piece was a type of
suzani called "large medallion." These typically are the priciest of all
suzanis, and figures above $20,000 are usual. This piece had the most common hexagon
medallion with four rosettes in the corner, and it sold on the phone for $18,400. The
estimates on the Sofer pieces were generally below market because the consignor wanted to
sell, rather than reflecting realistic market value.
Sotheby's had also managed to snare two of the rarest and most desired Salor Turkmen
weavings for this sale. Sotheby's had sold two of these pieces in the December 1993
Thompson sale; one, a two-medallion weaving in excellent condition, sold for $46,000, the
other, a three-medallion piece with worn silk areas and losses to the sides, sold for
$51,750. Lot 82 in this sale was a three-medallion wedding trapping and was perhaps in the
best condition of the three-medallion type to be offered on the auction market. Although
the original edges had been replaced, the pile was unworn. The $30,000/50,000 estimate,
given past performance of these pieces, was certainly reasonable, but Munich dealer
Eberhart Herrmann was the consistent buyer of these pieces, and he is no longer active in
the market. The bidding opened rather weakly but proceeded steadily up to $46,000; it went
to a California collection.
The second piece, lot 117, was an even rarer Salor ensi. In November 1989 Christie's
sold an ensi of this type for $104,000, and a year later Rippon-Boswell sold another for
$135,000. This was going to be one of those tests of the market, to see if top pieces
could still bring top prices. But what is top these days? This ensi was in the best
condition of all known. The pile was hardly worn, it had its original selvages with only
the loss of the kilim end, and there was one small reweave. But compared to the other two
auctioned pieces and one exhibited in London in 1983, this ensi was uninspired. The colors
were too dark, and it needed more of the light and mid-blues. It needed to be hung on a
yurt entrance in a sandstorm to abrade its surface and give it some character. Given the
$80,000/120,000 estimate, the consignor had provided leeway for the market to have
retreated on these pieces. But rarity, not beauty, drives the Turkmen collector. The
question was, were there interested buyers? The bidding was slow and mostly to the phone,
which took it at $90,500, still a top price for a Turkmen rug but far below the highs of a
few years ago.
A Chodor trapping, lot 133, showed what can happen when estimates are unreasonably
high. This hexagonally shaped rug is apparently unique, at least no other Turkmen weaving
in this shape has been published. Its function is certainly not known. Although
interesting, it was not beautiful, and the $18,000/22,000 estimate was inflated, even for
a unique piece. Four or five potential buyers of this rug were in the audience, and not a
one bid, so it passed at $11,000. If this textile had been estimated at $6000/10,000, it
is likely that bidders would have participated, driving it up to its competitive level.
The high estimate, however, simply eliminated free-market forces by being too high.
Lot 90, a Moghan Kazak prayer rug, was another example of overestimation. Since this
rug was missing all borders except the innermost one, it might more appropriately have
been called a fragment. At least 4 inches to 6 inches were missing on all sides of the 4
feet 6 inches x 2 feet 4 inches rug (est. $10,000/15,000). It was passed at $5500. This
rug was published in Prayer Rugs from Private Collections (1974), and one has the
sense here that the owner feels because it was published it was worth "a lot of
money." Not so.
A more beautiful and popular rug was lot 136. This Eagle Kazak was in very good
condition and had a bold single medallion with archaic dragon elements at each end. In
today's market the $10,000/15,000 estimate seemed high, but this design remains popular
with the public in general and not just collectors, so after spirited bidding, it sold for
a deserved $13,800.
Many previewers may have missed a Turkmen main carpet that arrived for display two days
before the auction. This rug is one of the earliest Turkmen main carpets and deserves an
18th-century date. The closest analog to it is the famous Ballard carpet in the
Metropolitan Museum of Art. The field patterns consist of archaic forms that seem to be
the precursors of some 19th-century Turkmen guls. This rug was to be sold in Sotheby's
London sale on April 26 and was brought to New York for exposure. It may well have paid
off. Estimated at 12,000/15,000, it sold to an American for 32,000 ($55,000). But when
comparing the importance of this rug to Turkmen rug studies, its rarity, and the number of
collectors who don't have this type, it should have sold in the $100,000 range.
The gross total for this sale was $2,399,093, and 76of the lots sold. Sotheby's next
major Oriental rug sale is scheduled for September 13.