Long live King Hussain

CSFBÆs King of the spreads, Abdul Hussain wins our spread prediction competition.

Back in May we asked investors and fixed income research heads to predict where three bonds would trade by September 2. To entice investor's interest we offered a magnum of champagne for the investor whose predictions came closest.

Behind the game was an ulterior motive. We wanted to see who would perform better: the investors, or the credit analysts from the investment banks. We ended up with eight investors and eight credit analysts making predictions on the spreads for the KDB 06, the Petronas 09and the PLDT 12.

As it turned out, the bond that separated the men from the boys was PLDT's which widened out to 864.5bp over treasuries by September 2.

We calculated the winners by averaging the margin of error on their predictions for the three bonds. As can be seen, the overall winner was Abdul Hussain of CSFB, and in second place was Colin Teichholtz of Goldman Sachs Asset Management. Teichholtz is the winner of the champagne, since his was the best performance from an investor.

Top 5 forecasters

1 Abdul Hussain, CSFB
Average margin of error = 12.36%

2 Colin Teichholtz, GS Asset Management
Average margin of error = 14.29%

3 Fan Jiang, Goldman Sachs
Average margin of error = 15.16%

4 Nuj Chiaranussati, Deutsche Bank
Average margin of error = 15.25%

5 John Woods, HSBC
Average margin of error = 15.36%

As can be seen, the top five is dominated by credit analysts rather than investors. Indeed, the average margin of error for the analyst community was 18.86% versus 25% for the investors. Sadly, this tends to suggest analysts may be worth their salaries after all.

To give some idea of the ranges, the worst overall performance for an individual credit analyst was a 28.8% average margin of error, while the worst from an investor was 48.85%.

The investors also produced some of the most off-the-mark calls on individual positions. For example, two investors (one from a very large firm) managed to clock up margins of error of 61.1% and 44.44% on KDB bonds - which is a remarkable achievement for a bond that trades at a spread of 90bp over treasuries.

On the flipside, the two closest calls on PLDT came from investors, with the best call being off by only 13.24% (versus the worst call from someone who was off by a substantial margin by predicting PLDT would trade at 185bp over).

To calculate all of this, we used spread data from Barclays Capital - which did not participate in the competition. The final midpoint spreads used were: KDB +90bp, Petronas +182.5bp, and PLDT +864.5bp.