Understanding our position in Investec Value

Portfolio Review:

The poor performance of the first quarter of 2015 continued into the second quarter, with the portfolio lagging the benchmark All Share Index’s return by around 6%. While the overall market was more subdued over the quarter, the best performing shares were still the most highly rated stocks, albeit different shares from the first quarter. Strong performances came from Naspers (+4%), MTN (+12%), Woolworths (+16%) and Brait (+40%) – all shares we believe are significantly over priced. Once again the laggards of the past 3 years underperformed, with the major holdings in the portfolio all continuing to fall (Goldfields -27%, ArcelorMittal -29%, Sappi -11%, AngloGold -6% and Amplats -4%). The funds offshore holding via the Investec Global Value Fund rose 6% in both US dollars and rand and this contributed meaningfully to performance.

Portfolio Activity:

Over the quarter we switched Goldfields and Sibanye into Impala Platinum and sold out of Cashbuild entirely. We also added a small positive in Tiger Brands into significant price weakness.

Portfolio Positioning:

The US Federal Reserve cutting interest rates from 6% to 1% from 2000 to 2004 to re-inflate the economy post the dot-com bust created the US housing bubble and the 2008 global financial crises. If three years of between 1% and 2% interest rates created the US housing bubble, what has the last seven years of zero US interest rates created? The product of this manipulation of the world’s risk free rate is zero yield German bonds, the Chinese stock market bubble, US stock market trading on a 20 PE on double its long term profit margins, a bubble in social media stocks (Naspers trades on 70 times earnings and 12 times revenue) and a 19 PE for the South African stock market (with bond yields at 8.3% and 1.5% GDP growth – an equation that simply doesn’t work).

Seven years of a zero fed funds rate has inflated assets everywhere and has “front loaded” returns so that future returns from all assets are meaningfully reduced. While asset values have inflated, zero interest rates and QE has barely moved the needle with respect to the target that they were intended for – the underlying economy.

Seven years on and global growth remains anaemic with best performing economy (the US) only currently running at just 2% GDP growth.

While commentators point to the recent strong employment numbers in the US as a sign that all is well, we point out that these numbers are inflated by service industry jobs (bartenders and waiters) which have boomed – people who are employed to service the booming financial services industry inflated by QE?  The fact is that since the global financial crisis, 1.3m bartenders and waiters have been added to the US economy while 1.4m jobs have been lost in manufacturing.

The net result of this is that we find almost nothing to buy in the South African equity market and limited options offshore. The Investec Value Fund’s allocation is thus 50% SA equity (mainly gold and platinum shares), 30% offshore equities (mainly Japanese equities) and 20% cash (effected via a short position in both the ALSI 40 and the S&P500).

Our gold position (AngloGold and Goldfields) is predicated on our view that the upcoming slowdown in the US economy will result in the US eventually having to follow the Japanese and Europeans with more QE – the run-up in the gold price from $800 to $1900 after the global financial crisis was prompted by QE globally.  In addition, gold shares trade at 20 year low valuations.  Our platinum position (Implats, Amplats, Lonmin and Aquarius) is a result of 20 year low valuations and the fact that platinum group metals are the only major metals currently in deficit. Outside of gold and platinum shares, we also hold Sappi (forward PE of 11 and continuing to deleverage) and ArcelorMittal (market capitalisation just 5% of replacement cost and the cheapest steel company in the world on an EV / ton of steel produced basis).

A final point on risk – Investec Value focuses our appraisal of risk entirely on the company and the industry concerned, and see no value in trying to appraise risk via the analysis of what others are doing (“tracking error”) or what historic returns and volatility are (“standard deviation”).

The examples below illustrate the above mentioned points:

30 June 2008
	Price ( R )	Volatility (SD %)			Returns (%)							Weighting in
		1yr	3yr	5yr	1yr	3yr	5yr	Price to Book	Price to Sales	PE	DY (%)	ALSI
Anglo American	550	30	23	26	30	239	359	4,0	3,9	16,0	1,8	17,2
Mr Price	14	31	30	30	-14	26	173	2,5	0,5	7,0	6,5	0,1

June 2008 marked the market’s peak optimism with respect to the commodity “super cycle”.  Oil was $140 per barrel and was expected to go $200 and as a result domestic retailers were out of favour.  Anglo American was R550 and Mr Price R14 and looking at traditional measures of risk, Anglos appeared lower risk. Volatility (measured as S.D.) was lower than Mr Price and trailing returns between 1 and 5 years way superior to Mr Price. Using tracking error as a measure of risk would have resulted in a “low risk” portfolio allocating up to 17% of the portfolio in Anglos and just 0.1% in Mr Price.  However, using valuation as a guide to future returns rather than backward looking risk measures would have resulted in a much bigger allocation to Mr Price than Anglos in 2008.

The gap in valuation was enormous in June 2008 – Anglos traded on a 16 PE (on record earnings) and four times book and sales, while Mr Price (an annuity earning company) traded on a 7 PE, 0.5 times sales and 2.5 times book.  In June 2008, Investec Value fund held no Anglos and 3% in Mr Price.

Anglo American only went up with low volatility.  Fast forward seven years later and the numbers look like this:

20 June 2015
	Price ( R )	Volatility (SD %)				Returns (%)								Weighting in
		1yr	3yr	5yr	7yr	1yr	3yr	5yr	7yr	Price to Book	Price to Sales	PE	DY (%)	ALSI
Anglo American	179	30	29	28	35	-27%	-26%	-35	-65	0,8	1,0	16	5,8	3,6
Mr Price	240	20	24	24	25	42	132	277	1510	12,2	3,5	26	2,4	0,8

As shown above, seven years later Anglo American is 65% lower and Mr Price is 16 times higher and now Anglo’s looks like the high risk company with negative returns over all time periods and volatility now well above Mr Price level.  Using tracking error as a measure of risk would result in a 3.6% allocation to Anglos at R179 versus a 17% allocation at R550 seven years ago.

Fundamental valuations now show that Anglos is low risk and Mr Price is high risk – the opposite of the market’s perception and what tracking error and S.D. of volatility tells us.

The table below gives another pertinent example in today’s market:

	Price ( R )	Volatility (SD %)			Returns (%)							Weighting in
		1yr	3yr	5yr	1yr	3yr	5yr	Price to Book	Price to Sales	PE	DY	ALSI
Naspers	1900	29	26	27	54	314	464	10,0	11,2	70	0,2%	9,9
AngloGold	110	58	49	41	-36	-58	-58	1,3	1,5	16	0%	0,6

Traditional risk measures show Naspers as low risk and AngloGold as high risk (Naspers’ five year return is 464% at 27% volatility while AngloGold’s five year return is -58% at 41% volatility).  A zero tracking error portfolio would allocate 10% in Naspers and 0.6% in AngloGold.  Naspers trades at 10 times book, 11 time’s sales and a 70 PE – AngloGold 1.3 time’s book, 1.3 times sales and 16 PE.  Investec Value currently holds 13% in AngloGold and no Naspers.

In conclusion, despite nearly four years of underperformance, we remain committed to our long standing views and believe in time our positions will be vindicated.

Please let me know if this helps or if I can provide further information and thank you for your continued support.

Regards,

Rod Hunter