What a week! Just a few days after I wrote a blog about buying silver it crashed! The blood is on the streets, on trading desks, on margin accounts and everywhere else it can be. OK, lyrics aside - what has happened and what is likely to happen next?
Doing a simple analysis of front contract daily price changes revealed that there were 12 similar or worse crashes during last 50 years. And here is crashes hit parade and the events that happened to take place on those lucky dates:
22-01-80 -22.73% Hunts times
26-03-80 -21.78% Hunts times
23-09-11 -17.75% Recession fears
14-03-80 -17.65% Hunts times
28-02-83 -17.60% n.a
17-03-80 -17.14% Hunts times
20-04-06 -13.75% Margins changed
13-06-06 -13.01% n.a
13-03-80 -12.97% Hunts times
02-10-08 -12.92% October 2008
20-10-87 -12.02% Black Monday
17-09-81 -11.30% n.a
5 times out of 12 drops took place when Hunt brothers attempted to corner the market,
3 times drops happened when all markets dropped,
1 time it happened when margins were hiked but I could not identify events that caused that, and
3 times I could not identify anything at all.
Speaking in terms of VAR (value at risk) the above events represent 0.1% of total population (12 out of 12500 trading days). VAR analysis takes into account 95% or 99% confidence level not 99.9%. In other words what happened on 15th cannot be explained using one of the most used approaches in today's risk management and it cannot be predicted. It just happened - markets are not perfect.
And since we are dealing with something very unusual let us see how prices reacted in the past. I am going to drop "Hunts times" events from further analysis as those are not relevant - we are not in a situation when someone artificially inflated prices of a single commodity, the whole metals sector and many other markets took pain on 15th:
23-09-11 Having finished the day at $30 level it took markets 2 months to hit $35 again
28-02-83 It took 2.5 months to recover from $10.3 to above $13 levels
20-04-06 In 3 weeks prices jumped from $12.5 to $14.8
13-06-06 Again, in 3 weeks prices appreciated from $9.6 to $11.3
02-10-08 This time it took 4 month to get from $11.1 to $13.1 levels
20-10-87 Prices attempted to bounce but stuck in the range and in several months a long-term silver bear market began.
17-09-81 Another bear market that lasted for 1 year
5 times out of 7 prices recovered quite fast but there is no unanimity.
Let's try another voodoo spell - maybe it helps? This one is called Commitments of Traders report. Let's see positioning of users of silver and speculators during periods of price crashes:
23-09-11 Users - Ultra Bullish, Speculators - Ultra Bearish
20-04-06 Users - Neutral, Speculators - Neutral
13-06-06 Users - Bullish, Speculators - Bearish
02-10-08 Users - Ultra Bullish, Speculators - Ultra Bearish
20-10-87 Users - Ultra Bullish, Speculators - Ultra Bearish
2 times out of 7 data was not available. Other 5 times users of silver never had bearish stance. In fact 4 times they were feeling the bottom and they were buying bottoms. No wonder they are called "smart money". And today users of silver are extremely bullish again. They have been bullish since early Feb, they realized market bottoming. But of course they did not and could not predict what happened on 15th. As I say above this drop had a chance of 0.1%.
Speaking of bottoming - how do users of a commodity realize it? Well, here comes next portion of black magic - let's see how much it costs in the real world. Take an example of Hecla Mining Company, largest US silver miner. Download their audited form 10K for 2012 and do some homework. What we are looking for is a benchmark indicator for cost of silver - in real world no one would be selling it below its costs, right? Not in the long run at least. So, we are looking for:
(NYSE:A) - gross sales
(NYSE:B) - profit from operations adjusted by costs of one-off events (they are not relevant to usual production costs).
(NYSE:C) - cost of operations
(NYSE:D) - indication of silver dollar value using product sales during 2012 multiplied by average prices during the year
(NYSE:E) - share of silver dollar value in A
A = 321M
B = 38M + 25M (one-off event cost) = 63M
C = 321M - 63M = 258M
D = 6,4M SI oz * $31 = $198.4M
E = 198.4 / 321 = 62%
Hecla cost of silver production = C * E = $160M
Cost of 1 SI oz = $160M / 6.4M oz = $25
Hecla will not be selling silver below the cost for too long. They would rather close down the business. And by the way - they reduced silver output by 1/3 between 2011 and 2012. Guess why... Commodity markets react fast, another example - last April natural gas prices were below the cost at $1.90, today it is $4.10. And by the way - they state they do not hedge the price. Probably, they know that from this point prices have no other way to go but up.
OK, let's get to conclusions. The only time when prices did not recover well was the case of Black Monday. Why did it happen? Well, hard to say now but it do not think we are in that situation. In 1987 US Government tightened monetary policy and it was reasonable to sell assets and sit in cash. But today situation is very much different. Some say inflation is up to 10% p.a. In all other cases it took weeks or months to score a 2-digit recovery. What am I going to do? I will wait a few more days and if Black Monday won't be here I will buy again. And of course I will hedge my long bet buying puts. Just in case.