Page 4
(Page 4 Winspear)
Notably both Klipspringer and Snap Lake are "Group 2" kimberlites (as are
all of the other South Afnoan "fissure" deposiIs) whioh comrnonly have a
high proportion of large diamonds.
The market has focused on the high proportion of the total value of the Snap
Lake diamonds thax is carried by fhe 3 largcst stones.In it’s release
of January l5 Wirspear notes that those stones which make up slighdy more
than 10% of the weight recovered carries appaamakly 75% Of the totai vatue
of the 226 carat parcal (from Winspear's fIgures we infer a vaiue for those
diamonds of $2035/cara4). The average value of the remaining 201 carats is
given as $86.26/carat which we believe likely breaks down intp something
like $175-200/carat for rhe remaihing +t earat diamoads and $50-55/carat for
those diamonds less than 1 earat (compare this to an average value for
diamonds from the Diavic projeot of around $60/oariat).
We feel it is usefu to focus more closely on the valuation data for the Snap
Lake and how it gompares with other kimberlite deposits. As shown in thc
table beiow there are a few kimberlites with average values in excess of
$200/caraL but aI this point none are known with vdues exceeding $285. Again
it is worthwhile noting that most of the very high value kimberlites like
Snap Lake belong to Group 2.
Table - Parcel Values of other Group 2 Kimberlite Deposits
Deposit Value
(/carat)
Blellsbank (Rex} $285
Koffiefontein $211
Koldu $200+
Kudu (explortion sample} - Klipsprincer $203
Ardo $200+
Bellsbank (Messina) $185
M-l (Marsfonteln - Kiipspringer $170
Leopard - Kllpspringer $130
Star (Rex) $114
Sugarbird - KliDspringer $112
Source - Company Reports and DBS estimates
To look at the eoncenkrarion of value in the small nurnber of >1 carat
stones we hive fewer comparable data. Rex tDamonds (RXD-TSl~:$lAi) rerports
a general breakdown of revenues and values by diamond size from their South
African operations, indicating that about 77% of their revenue comes from
diarnonds over O.85 carats, which represents 32% of their production by
weight
It is generally accepted that the value of diamonds recovered from
exploration samples tend to underestimate the actual value of diamonds in
the deposit as a whole.This is because the scarcity of the larger ( and
rnore valuable) diamonds leads them to be underrepresented in
small sample (and we would consider 226 carats small). The flip side of this
is the danger that one (or a few) large diarnoads can bias the vaiuation
results upavards quite substantielly m a small sanple. Were the Winqpear
valuation dominated by a single, large, high-vaiue diamond we vould be
signifieantly more cautious. It is our experience, however, that "getting
lucky" with one diamond is possiblo, but that with three stones each oanying
a substantial part of tiie parccl’ s value the likeliihood that the
veey hiigh valuation obtained for the Snap Lake diamonds is gossly inflated
is, we believe small When er say this, it is still withi the feeling that,
in the finai anaiysis the vaiue may be oloser to S200 than $300/caraL
Possibie upside in grade (see below) may iargely compensate for a likely
lower value, however.
New Causti Fusion (Microdlamonf) Data
In our opinion, through WSP s ran-up to armost $5. 00, the big gap in the
story was the availability of data that allowed comparison beween the two
mini-bulk sample sites (Pits 1 and 2) and the much larger inferred
kimberlite body. The January 26 release provided this "rnissing linle", and
in our opinion dle data eould hardly have oeen more posidve The market
response has mostly been the opposite however, \vhkh we feel is mostly due
to the opaque nanare of the new data.
Dia mond Abundance
The data (see table) show that in terms of diamond abundance (as stones
normaiized to sample weight) the drill hole results are fully comparable
with rhose from Pits 1 and 2, and are in fact better. on average, better
even than the higher grade Pit 2. This applies to all areas, inrluding the
norihwest dike samples, thle dovn-dip drill holes, the north shore sampIes,
as well as the gaologically distinct Southeast dike IIt also appiies to
diamonds in all size classes, from the <015 mm to the >l .O8 mm. In rhis
connection we note that looking just at the data for the two pits. the total
diamond abundanee (oalculated as stones/ tonne) correlates very well wkh
grade
(continued .)
(Page 5, Winspear)
of th two samples, and that with a greater stone count in rhe driil core
samples the possibility of a grade of around 1.5 carats/ yonne is reai. This
may serve to comper sate should the diatnond values not be rnaintained at
the present estimate of $300/oarat
MacrorMiero Ratio
The other important variabIe in assessing the significance of caustic fusion
data, in adition to absolute stone abundance, is the ratio of larger to
smallar diamonds. Because most drill sarnpies are too smali to expect
ommercial-size diamonds to occur, this ratio provides critioal information.
Since the dimnond exploration boom began in 1991 it has been traditional in
Canada to report "micro" and "macro" diamonds, with macrodiamonds defined as
those writh a maximum dimension greater than 0.5 mm. In other Canadian
kimberlites thc the micro/macro ratio typically ranges between
2:1(verygoodl) to more than 10:1(not so good). ln this current release
Winipear has broken with this tradition, and reports diamond abundance in
terms of square mesh size (which is rnore defensible on technical grounds).
Looking ar the table it is clear that for the ratio of diarnnnds <0.30 mm to
those >0.30 but less than 1 08 mm, that the population represined in the pit
samples is similar to fhe drill core data. For the larger stones (.1.08 mm
there is more the scatter,in the data, but this is not unexpected, as the
odds of finding a +I mm stone in some of the very small samples (less than
40 kg) are smalL Nonetheless, if we take all the drill hole data together
and cornpare it to the Pit l and 2 data the proportion of larger diamonds is
actually somewhaú higher in the drill hole samples. All this comes together
to suggest that there is no reason to expecy any fewer large diamonds in the
dike as a whoie as oompared to the onginal mini bulk sarnples.
.
Commerrial-Size Diamonds in the S, plei
Looking closely at the recent Winspear release brings up another poin we see
as very favorable, and one that seems to have been mostly overlooked by the
rnarket. In the approximately 1.7 tornes of the northwest dike treated oy
caustic fusion and discussed in rgis retease. at least 6 diamonds larger
than 0.1 oarat (anld therefore certainly of commercial size) were recovered.
These include two diamonds larger than 0 5 carats (0.75 and 0.69 carats),
and have a total weight of almost 2.5 carats - therefore on these stones
alone the samples had an average grade of 1 45 caratsftonne, in line with
thc higher grades encoumered in Pit 2. We also rernind the reader that,
beeause of the log-normal distribution of apparent grades in kimberlite
siunples, smail samples (such as the drill cores) are more llkely to
understate, than oventate, diarnond grade.
Table 2 + Riauits of Caustic Fusbn Analysis, Snap Lake
STONE>.15<.30 STONE>30<1.08 STONE>1.08 RATIO RATIO
SAMPLE SAMPLE MM ST/10KG MM ST/1000KG NORMALIZED <.03 <1.08
WT
NORMALIZED NORMALIZED ST/1000KG l<1.08 :>1.08
PIT 1 262.0 19.2 484.0 26.7 3.97 18.1
PIT 2 274.0 36.5 879.0 7.3 4.15 20.4
1-11 53.8 39.3 744.0 37.2 5.15 20.0
J-11 368.4 40.8 983.0 57.0 4.15 17.2
J12 86.6 31.9 895.0 12.1 3.56 74.0
K-11 240.8 40.1 1042.0 41.5 3.85 26.1
K-12 88.1 37.4 1022.0 11.6 3.56 88.1
DETALED
AREA 369.1 35.0 864.0 24.3 4.17 35.6
NORTH SH 38.7 50.1 1136.0 0.0 4.41 N/M
BW DIKE
DD 33.4 33.0 810.0 30.0 4.-7 27.0
SE DIKE A 18.6 46.2 1130.0 0.0 4.0 N/M
SE DIKE B 42.4 41.5 1157.0 94.4 3.59 12.3
SE DIKE C 82.3 31.6 658.0 24.3 4.73 27.5
Source: DVS estimat and company reports
Share Impact and Evaluation
As should be clear from the above discussion, we consider the present
results very positive for Winspear, as they strongly suggest that the
ongirlal two mini-bulk samples were collected from sites representtaive of
lhe rest of the body. The new data make it clear that the diamond
populations in lhe mini-bulk aempies, end the 'oody as a whole are ompleteiy
compatible. The larger recovered during the caustic fusion yield a grade of
at least 1 .45 carats/tonne, similar to that in the higher grade of the two
original pits
(Page 6, Wmspear)
We therefore feel it is likely, WinKspear’s caution norwithstanding.,
tbat thls year’s plannod 6000 tonne bulk sannple will yield grades in
the 1.1carats/tonne range or possibly higher.The main uncertainty is rhar of
value - clearly ewaults similar tothose for the original 226 carats would be
very positive. As we discuss above , however, any value above $170/carat
should generate an acceptable return even assuming unit mining and
processing costs 50% higher than thos assumed in Winspear’s scoping
study, andabout double those realized at the nearest (former)operation, rge
Lupin Mine of Echo Bay (ECO - TSE:$2.79). This includes initial capex of $US
230MM for a 3000 tonne/day rate(vs the company’s figure of $63 MM for
1000 tonnes/day), and sustaining and expansion capex of over $110 MM life of
mine (vs. $3MM based on the company’s figures).
Table3-NAV’s and lRR for different Grade- Value Combinations Snap
Lakei
Reserves Grade Value Capex mm OperatingCOST NAV 8% NAV - 0% IRR
$CDN /SH $CDN/SH
14.5MM T 1.14 c/t $301 /ct $US 172.5 $US 49/T 17.35 43.00 51 %
14.5MM T 1.6 c/t $225 /ct $US 230 $US 73/T 14.50 37.30 41 %
14.5MM T 1.5 c/t $175 /ct $US 230 $US 73/T 9.30 25.25 32 %
14.5MM T 1.14 c/t $225 /ct $US 230 $US 73/T 8.90 24.30 31 %
14.5MM T 1.14 c/t $175 /ct $US 230 $US 73/T 4.93 16.13 23 %
Source: DBS estimates and Company reports.
NAV represent WSP’s approx imately 68% interest.
lf we accept the $301 /carat at faee vatue, along with operating costs in
line with the company’s estimates, we estimate an NPV for
Winspear’s share of the project of $780 MM CDN or $17.35 /share (using
an 8% discount rate). This is based on mining a total of 14.5 MM tonnes
through 2015. With $175 /carat and operating costs about 35% higher, our
estimated NAV falls to just under $5.00 /share.
If we look for a 2OO3 start-up land discount these NAV's baek to l999 at l5%
'we get a range of values between $3.70 and $10.75 for Winspear. On this
basis we see that, even at its highs Winspear’s share pnce has always
discounted the lower end of this range. Given the uncertaonties in the grade
of the kimberlite and the diamond value we feel that this was appropriate.
Hiowever, with the new data suggesting that the grade has a good chance of
being maintained we feel quite coumfortable witrh the current valuation .If
the results of thls wmter’s bulk sarnple program support a higher
value than $200 /carart then we see the potential for Winspear to appreciate
signifiriantliy. We do see some hurdies ahead for the cotnpany, however, as
the special nature of the deposit may make finarcing unusualiy diffcult.
Thr Trouble with Diamond Deposits
The difficulty we foresee fot Winspear (and partner Aber); is in
establishing an esttmate of mirieral inventory that will rneet the needs of
poiiible lenders. To drill out the kirnberlite and prove tonnage shouljd be
no problem - in fart easier than for many types of deposits. We suspect that
with additional data it may even bep possible to predict diamond grade
within acceptable limits based on caustic fusion data from drill hole
samples. The missing link could be the diarnond vaiue. We see no way for
Winspear to establish rge value of the diamonds (to a bank’s
satisfaction) without a substantial underground program ehich would sample
the body in several blocks. This has implications for timing development as
well as financing needs (and therefore dilution). We believe the earliest an
underground program could begin is early 2001, given rhe requirements of
permltting.Allowing for a 9 month program followed by 3 months of sample
processing,, the results would be iknown in early 2002. Such a program would
likely cost in the tegion of $20 -25 MM (Cdn) reqtliring WSP to fund
approxvaately $15 MM - money it would have to raise. To complete project
feasibility study, arranige financing. and obtain permits ii likely to +take
a minirnum of 2 years (bringing us to 2004). Assuming that everything fell
mto place to allow mobilization of equipment over the winter road durmg the
2003-2004 season the earliest we would forecast produetion (from an initlal
open pit) is early 2005
Other possible routes exist, however. Winspear might opt to begin
feaiibitity aind permitting for an initial small- scale (say 1,000
tonnes/day or less) open pit operatian (assumvng the current bulk sarnple
yields positive results). and nm the evaluation of the larger underground
portion of the deposit in paradlel with thwt effort. Whde this would not
rnaterrially shorten the lead time to a decision on developing the larger
portion of the deposit, it could yield cash flow during a critical period,
as well as provide additional confidence on the continuity and grade of the
sill. Even in this situation however, rgw company might not be self funding
(unless the very high diamond values are maintained), and so some dilution
of current shareholders might be unavoidable.
Conclusion
We have liked the potential of \Vinspear s Snap Lake propery since we first
saw the diamond grade and values about 6 months ago. The missing data to let
us fully assess the project was mainly the microdiamond results fiom the
mini bulk ssmpIes. We now have those data which strongly suggest that the
kimberlite in the rnini-balk samples was typical of the main northwest
dike,, and other dikes at Snap lake also contain similar populations of fine
diamonds.
In our opinion this indicates a good likelihood that similar (> 1
caradtarne), or possibly even better grades, will is found in the 6,000
tanne bulk sampling period planned for this winter. The main uncertainty
would then shift to diamond value - the previously obtained $301/carat is
among the highest valuation ever obtained, and we question whether such a
value could be maintained . As long as the value rernains over about
$170/carat we estimate that an acceptable return of investment is possible,
however, if a grade upside we suspect does materialize this threshold value
could drop to the $150/carat range. Other (Group 2) fissure kimberlite
depodsits in Africa do reach or exceed these values, so unprecedented high
values are not needed.
In the long term we do hsve a major concem that the time lag to production
and some of the unique challenges in financing this deposit could weigh on
the stock. Even if an ecoIlumic deposit does ernerge, Winspear may follow
the traditional trajectory of a rising share price during exploration and
delineationL, with a possibly long dull period leading up to productlon.
Rccommendation
Winspear has advanced Snap Lake to the point where the plzy has to be taken
seriously. The future is in the numbers - the grade from this vear's bulk
sample, and equallly important the diamand Vahues.. With grade and value
equating to about $185/tonne or greater we see a good probability of a mine.
We feel that the available data make such a value likely, Until these data
are in Wnspear remains a speculative stock, but h our opinion one nf the
best high-risk, higll-reward plays in the junior market. We rate Winspear
Outperform (Speeulative). |
