What a month! Since it was so awful (All Share down 13% for the month and 21% for the quarter), and since readers will already have been bombarded by gloomy news, let's start with two cheerful observations from September:

1. On September 29th, a day US stocks plunged the most in percentage terms since 1987 (and the Dow Jones fell by a memorable 777 points), only one of the 500 shares in the S&P500 was up: Campbell Soup. Maybe Americans were feeling like South Africans ahead of the 1994 elections and stock piling essentials! (Remember the run on candles and baked beans?)
2. In a quarter filled with terrible earnings announcements from corporations all around the world, there was one fabulous exception: SSL International plc, manufacturers of Durex condoms, reported an outstanding year.

Now to a quick chronological recap of just a few of the momentous news releases in the month:
Mortgage giants Freddie Mac and Fannie Mae were taken into conservatorship (a euphemism for being nationalised) on Sunday 9th. Together, they have$5.4 trillion in liabilities. The US government will inject $100 billion of new capital. Both CEOs were promptly fired.

Lehman Brothers, an icon of Wall Street, filed for bankruptcy on the 15th. On the same day, Merrill Lynch was obliged to succumb to a take-over by Bank of America.

AIG, once the world's largest insurance companies, saw its share price fall 75% on the 16th and the Fed created an $85 billion credit facility in exchange for an 80% shareholding. So, another nationalisation, in effect.

In the UK, HBOS (itself a result of a 2001 merger between Halifax and Bank of Scotland) was, under duress, acquired by Lloyds TSB.

Amid many other collapses and imminent collapses, the mother of all bail-outs was announced by US Treasury secretary Henry Paulson: TARP (Troubled Asset Relief Program), a $700 billion plus package. The markets staged a short-lived and ill-fated rally because later, the House of Representatives voted it down, sending markets into a tailspin. The vote appears to have had more to do with self-serving, individualistic politicians (is there another kind?) with an eye on impending elections than a sense of what was collectively required. The indefatigable Paulson revised the proposal to include the same essence, but to also pander to popular requirements such as caps on executive remuneration. We note two other bizarre inclusions: (i) a requirement that health insurance companies provide more coverage for mental health services; and (ii) an excise tax exemption for producers and importers of certain types of wooden arrows used by children! Well, whatever it took, it worked and the Senate approved the modified bill on Wed 2nd Oct. At the time of writing, the House of Representatives had to agree it too, but there seems little doubt that it will pass this time.

While shares were falling fast, several governments panicked and banned or severely restricted short selling (Australia, Netherlands, UK, USA and others).Back in the Asian crisis, the bewildered Malaysian Prime Minster also banned "evil" short sellers, and much of the "first world" condemned him for interfering with free markets. Even further back, privatisation was viewed as a hallmark of advanced capitalism, while nationalisation was the scourge of communists and socialists. What a turn for the books! We are adherents of the theory of mean reversion, but we could hardly have expected to witness it on this scale. In a moment that swelled our patriotic pride, the JSE announced that it felt no need to resort to such drastic measures.

So, how bad was September really? Well it was quite bad, but not record setting. Looking at monthly data on the S&P500 from January 1871 (so we have1,654 months of data), we find that September (-14%) was the 11th worst single month. The worst was November 1929 (-26%). The extent of the 1929 crash (the nadir occurred in June 1932) was 84%! The current drawdown, starting in October 2007, is a measly 28%.

And how bad was September for the JSE? We have data since December 1960, which again shows that at -13% September was pretty bad, and was the10th worst single month. August 1998 (not, as popularly believed, October 1987) was the worst single month, at -29%. The worst draw down was between April 1969 and October 1971, a period in which the JSE fell 62%. The current draw down, which started in May 2008, is 25%.

So much for history. Where are we now? Firstly, the massive write-downs that have occurred globally ($500 billion so far, which have resulted in an astounding $19 trillion loss in market capitalisation, for a ratio of 38 to 1) simply have not been a feature of South African financial companies. Yet, as we have seen, financial market contagion has been fairly savage. This has rendered the PE (price: earnings multiple) multiple on local banks an attractive 7.9 times and the dividend yield an equally attractive 5.1%. Perceptions of a global slowdown have had particularly deleterious effects on share prices of locally listed resource shares. The RESI20 has lost 38% in the past quarter. As an aside, the same fears, coupled with absurd starting valuations and general hubris, have come home to roost in an even bigger way for the vaunted so-called BRICS. Current drawdowns are: China (61%), Brazil (42%), India (47%) and Russia (57%).

The fact is that South African banks have not been exposed to the same risks as have presaged the current crisis, and in fact have been doing a good job by global standards (among the best real return on equity, higher interest margin than UK or European banks, and well above average solvency ratios).

We'll try to end this on a positive note: recent events have rendered the local equity market cheap. Going back to the start of the "New South Africa", the PE of the JSE has only twice been cheaper than now: the aftermath of the bursting of the technology bubble and again in 2003. No doubt you are aware of the fantastic returns that JSE investors enjoyed in the ensuing four years. On a more scholastic note, we have rigorously tested the phenomenon of buying when markets are cheap (measured on this PE ratio), in South Africa since 1973 and in the USA since 1871. The results are emphatic: although cheap markets are always associated (by definition, we would suggest) with heightened levels of pessimism, projections of financial Armageddon in the popular press and pathetic posturing by at least some politicians, they are the best times to buy shares. The adjunct to this finding is that it requires great mettle to remain dispassionate among such widespread misery. We have the requisite intestinal fortitude to do so and we think that those of you who share this view will be well rewarded in time. To re-iterate, we don't know when the bottom of the current market cycle will occur, but we do know that the market is cheap and that it will recover.

In a month of such momentous news for equities, it hardly seems worth mentioning bonds, other than to say that fait value has become even more stretched after yields fell moderately, even as both consumer and producer inflation rates were once again higher than the previous month and higher than expectations. We estimate that for bonds to represent reasonable value, inflation must average below 4% over the next 10 years; an extremely tall order from a starting point of over 13% for CPI and over 19% for PPI.

Eventual acceptance of hard facts that were previously delusionally denied can be painful. So it was for the bond market in May and June, when 10-year yields rose 125 basis points, the All Bond Index lost 4.2% and the Long Bond Index lost 7.4%.

For as long as we have been managing this fund (since July 2007), the bond market has strangely chosen to believe that both increasingly higher rates of announced historic inflation and a worsening outlook for future inflation, were two problems that would miraculously disappear soon. To statisticians, economists' overwhelming concentration on the 12-month change in the consumer price index (as opposed to some other time frame) appears rather arbitrary, but this is an entrenched practice. So, the market is fond of pinning its hopes on the so-called base effect - the change in the inflation rate that is attributable not to the most recent data, but to the dropping out of 13-month old data. Sometimes this is valid analysis, as in 2002 when the rand had already recovered from its precipitous fall in 2001 and it was clear enough that CPI would follow. More recently though, benign base effects are less obvious: the oil price has continued to rise, electricity prices are sharply up and set to continue rising for the foreseeable future and wage settlements are mostly north of 10%. Yet bond yields have remained stubbornly below the contemporaneous inflation rate. Our scepticism and consequent short duration position has paid huge dividends within the fixed interest portion of the fund. We estimate the bond portion's performance to be 400 basis points ahead of the All Bond Index.

Equities had a poor month, save for resources, which just about managed a positive return and are up 33% year-to-date, while industrials and financials fell about 10% in June. The fund's equities mercifully managed to lose less money than the market or most of its competitors.

The massively demanding rating on MTN came to haunt the share in June, as the twin necessities of further good operational news, over and above the dazzling future already discounted, and favourable corporate action both stumbled. Meanwhile the much-maligned Telkom, which we hold in preference to MTN (in stark contrast to the great majority of domestic fund managers), hardly had good news. It simply enjoyed a cessation of bad news and a hint of positive news. Since their relative fortunes changed modestly, Telkom has outperformed MTN by 48%. In our view, it remains comfortably the cheaper of the two shares.

We have trawled back through data as long as we can, and find the dispersion of first tier sector ratings to be larger than ever before. Notably, financials have never been as relatively cheap compared to the rest of the market as currently. Granted, there are significant head winds for banks, but these companies' collective historic performance is a fact. Over 33 years, which included what were much worse episodes and outlooks for their earnings than is currently the case, they have, with some consistency, outperformed the overall market in the contest of earnings growth. They are now at their lowest ever price: dividend level relative to the market. As always, we have no idea when this anomaly will unwind, but we are comfortable that it will and that when it does, there could be very positive returns from shareholdings in banks.

During March the fund deployed15% into the Nedgroup Investments Global Balanced Fund managed by Sarasin Chiswell who are based in London and have an enviable track record. The fund retains its Regulation 28 compliant status, meaning that it is an eligible vehicle for pension funds. Motivations for this transaction included the observation that the JSE's P:E ratio is now at a 30% premium to that of the UK FTSE100 and is in close proximity to that of the USA's S&P500. This after the JSE has outperformed Wall Street by an astounding 263% this millennium (both markets measured in US dollars). From these starting valuations, the JSE has historically shown a strong tendency to underperform those markets when measured in common currency. Further, the JSE's dividend yield is at an all time low multiple of the Japanese dividend yield. Diversification is the paramount rule of investments and given current relative valuations, we expect this transac to be rewarding.

Fundamentals for domestic bonds continued to worsen in March. February's CPIX came in at 9.4%, embarrassingly far above the official target of 3% to 6%. PPI came in even higher at 10.4%. Pending electricity and petrol price hikes will not help, but bond yields imply that inflation is heading inexorably lower soon. The fund was well positioned; it has a very low weighting in bonds, and within that segment, the modified duration is also low.

The JSE All Share index lost 3% in March. Financials performed particularly poorly, losing 6.6%. This was not helpful to the fund. Conversely, it affords the fund opportunities to acquire some of South Africa's most solid financial institutions at the cheapest valuations we have seen in years.

Fundamentals for bonds and the rand worsened in the month, but both markets weakened only by almost imperceptible amounts. The most notable data release in December was the latest current account deficit, which is now a massive 8.1 % of GDP. That, combined with inflation figures that have been higher than economists' expectations and the Reserve Bank's targets, places a burden of very optimistic future economic outcomes on the current level of bond yields. We do not much like investments that are beholden to favourable yet largely unpredictable prospects. So, the fund retains a light weighting in bonds and rather meets its regulatory requirements by enjoying a 300 basis point yield enhancement in twice as much cash.

We would have thought that the madness of bubble prices would not occur more than once in an investment generation, which is probably about 15 years long. Yet, less than 10 years after the infamous TMT bubble, some strange transactions have occurred recently. Microsoft leaves us dumbfounded as to why it paid 100 times revenue for a very small stake in Facebook. At home Naspers, itself an unhappy participant in the TMT saga, this month bought a little known London domiciled internet business, active in eastern Europe called Tradus pic at a rating of 20 times sales, 63 times operating profit and 16 times NAV. For the price paid, Naspers could have bought the entire share capital of any of Medi Clinic, Afrox, Highveld, Santam, Woolies, Truworths, Foschini, Massmart, Nampak or Didata. We listened in on management's conference call with investors and were reminded of the fuzzy statements that used to accompany acquisitions when TMT was all the rage. Just one analyst queried the price paid and management's reply seemed vague to us: they do not forecast earnings, but they feel certain, based on cash flows, that they will be vindicated. Interestingly, the market was efficient and it promptly eradicated 10% of Naspers's market capitalisation. This portfolio does not hold Naspers. As a reminder of insanity, the Tradus share price has moved from £5 to £18 (the price Naspers is paying) this year, but its all time high was over £350!

Other things that do not seem to change include the futility of most rating agencies' work. Globally they have heaped scorn upon themselves by downgrading many sub-prime related securities after the event. Locally, Moody's Investors Service downgraded Imperial Holdings after the company had announced expected declines in headline earnings. One could pay for these kinds of services or just watch share prices for free and get more prescient information.

The fund's equity positions continue to reflect our conviction (represented by relatively few holdings), fair degree of agnosticism towards index or peer group benchmarks (represented by substantial differences in holdings) and value philosophy.

The fund's most positive share in the month was its 10% (of equities) holding in
Amplats (up 11%) while the worst performer was its 2% holding in Oceana (down 17%).

As at the time of writing, we have planned some activity for the fund, but given
the seasonal illiquidity of the market, this will be implemented early in the New Year.