The performance cycle

Performance cycles not only work at all time degrees but also challenge the idea of hedging

We were the first to talk about long India and short China on February 2009. We were also the first to publish long Nikkei-short BVSP Brazil, long Russia – short Nikkei, long India – short Nikkei and long Nikkei – short China in a paper on performance cycles co-authored by Ionut Nistor and this author, published in the Kyoto University Journal in March 2009. We gave two time frames for pair performance.

One was multiple months and one performance cycle extended for years till 2013-2015. We also said that the Goldman BRIC Model was broken if BRIC country performance with the Japanese region was polarised. All the pairs registered gains and were up 27 per cent (long Nikkei, short BVSP Brazil), up 14 per cent (long Russia, short Nikkei), up 51 per cent (long India Sensex, short Nikkei), up 59 per cent (long Nikkei, short Shanghai composite). We have enclosed the paper as a weblink here.

Illusion and reality?
It was kind of shocking both for the academicians and for the capital market participants. A few even thought of such work as unworkable strategies. These were a few remarks posted below the article in Business Standard.

Suki – “How would you rate the precision high quality engineering skills in India vs that of China? Who is going to be the better supplier in the future? Are Indians less quality conscious than China?”

Nano – Money is a king at this difficult economic time, and China has more than $1.3 trillion in reserve and India none. The Chinese consumer market is much bigger than India’s.”

John – Talking about a great future does not bring a great future, no matter which “model” you used in your predictions.

We tried explaining that performance cycles work in both directions in favour and against China. This is what we said “Performance cycles are mathematical. The comparison is quantitative, which can have qualitative reasons. A few years back, it was long China – short India. The qualitative reasons could have been different in that case. If population and consumption was key, there would have been no East-West dominance cycles.

East was historically more populated then the West. Consumption is a creation of modern economics. Population and consumption or anything else cannot change performance cyclicality. Any small country with 1/100 the population of China can outperform China in percentage terms for a certain period.”

Click here to view detailed story

Performance and time cycles
The idea of a preordained cyclicality of performance working in markets challenges many economic think tanks dedicated to researching and forecasting the same. Performance cycles based on time fractals simplifies too much for the linear time generation of researchers to accept. Our idea is still not about dissuading researchers from conventional research but embracing something so repetitive and structured as time fractals also known as time cycles.

The cyclicality in markets between regional indices is a new thought even for macro funds playing between regions. How many funds do you know doing pair trading strategies between Japan and the BRIC countries? And how many investors do you know putting money in such funds? Even if there are funds doing this there is still little credit given to time and its cyclicality.

There is another reason why the idea escapes the majority? The one who understands performance also consider it as just one cycle not two, three or many. Human mind is predisposed to linearity. This is why psychologists call humans as shortsighted. There was no way behaviorologists would have called us myopic, if we could see cycles. Our inability to see cycles is at the core of behavioural finance and psychological errors.

We see only one up part of the cycle not two. We comprehend performance more than underperformance. We buy winners and sell losers. If we would have comprehended that today’s losers are tomorrow’s winners, we may be selling winners and buying losers. Above this we rarely look at both winners and losers together. For the majority today performance is secular not cyclical.

This is why investors fail to comprehend how long India – short China and short China – long India can make money. For example, take the HDFC Bank and ICICI Bank pair. One may believe HDFC Bank to be a sector leader, a notch ahead of ICICI Bank. With this information in the background a strategy of short HDFC – long ICICI BANK may be ridiculous. How can be buy and sell two highly correlated assets? We assume statistical correlation above time.

We don’t realise that correlation like everything else is cyclical and changes with time. Take two time series and compare them over multiple time frames and you will see differing correlation values. Beta, the highly touted hedge ratio suffers from the same problem. Take beta values for various time frames on two data series and you will see fluctuating values. Majority of our investment problems do exist because we don’t want to admit cyclicality, which exists everywhere and at all levels of time.

Long India, short China
The over discussed long India and short China pair can validate this. The pair made 55 per cent since we mentioned about it early this year. The maximum return that could have been captured between the two indices was 60 per cent in 83 calendar days during March 9, 2009 – May 31, 2009. One had to invert the pair from May 31, 2009 till July 12, 2009 for 90 days, Shanghai composite outperformed BSE Sensex by 21 per cent. From July 12 to August 30, it was long India and short China again for 48 days delivering 22 per cent.

This was a classic example how Chinese and Indian performance was shifting in a cycle from one to the other. We can illustrate this same performance cycle on larger time frames. July 31, 1994 till June 3, 2001 (seven years) China outperformed India by 100 per cent (annualised 14 per cent). June 3, 2001 till March 31, 2006, the performance cycle shifted (five years) in favour of India which outperformed China by 279 per cent (annualised 55 per cent). We can go to a weekly cycle or daily performance cycle and show the same shifting performance cycles between the two regions. Performance is cyclical and predictable.

Performance cycles challenge the idea of a hedge like nothing else. We talked about it briefly in our India H2 2009 outlook. Any pair which delivers 20-30 per cent annualised returns is an opportunity to invest and not to hedge. If strange sounding pairs between regions can deliver such returns, a hedging exercise built around avoiding a loss or gain is counterintuitive. No wonder hedging is a cumbersome exercise when not based above the underlying cycle.

We also have beta tracking techniques, which is constant adjusting of two legs to keep them of similar value. We calculated the beta of the BRIC country indices and Japan with the MSCI world index. Nikkei and Bovespa had the closest betas compared to the world index. Close betas warranted a value hedging and not beta hedging. This means that the Nikkei – Brazil pair also delivered 18 per cent annualised. Performance cycles make a strong case against conventional hedging. This is no surprise. Hedging without an understanding of time fractals is an illusion.

The author is CEO, Orpheus Capitals, a global alternative research firm

(With additional contribution from Anna Maria Michesan)