SYSTEMATICALLY TRADING GOVERNMENT BONDS USING SIGNALS FROM THE OVERNIGHT INDEXED SWAP MARKET

STATISTICALLY SIGNIFICANT AVERAGE PROFIT OF 1BP PER TRADE; BID-OFFER SPREADS COULD BE AN ISSUE*

A theoretical case is often made for the derivatives market responding faster than the cash segment to economic data and policy announcements. If this assertion is true, then signals from the synthetic space can be used to trade cash instruments.

In the Indian context, I tested this hypothesis using 5-year/10-year government security yields and the 5-year overnight indexed swap rates.

The results are encouraging. OIS rates lead the moves in the government securities market. There is evidence of a positive follow through (positive moves followed by positive, negative followed by negative). A simple trading rule designed to trade government securities generates statistically significant profits.

The back-testing exercise indicates average profits of about 1 basis point (t-statistic >5) per round trip. I have not accounted for bid-offer spreads. Yet, market makers can certainly exploit this empirical regularity (because they do not have to cross bid-offers).

Rule # 1: 5-year OIS and 5-year Gsec

Data: Daily closing yields from February 2, 2003 to April 29, 2009

Cross-Correlation Structure Suggests Positive Follow-through from Synthetic to Cash

Correlation (OIS (-1), Gsec (0)) = 0.22
Correlation (OIS (0), Gsec (0)) =0.50
Correlation (OIS (0), Gsec (-1)) = 0.01

Trading Rule

If Delta (5-year OIS Rate) > 0, then short 5-year government bonds
If Delta (5-year OIS Rate) <0,then buy 5-year government bonds
Delta: Daily close to close
Trade: Daily close to close

Performance Statistics*

Mean (bp): 1.1
Standard deviation (bp): 6.7
t-statistic: 6.6
No. of Positions: 1498
No. of loss-making positions: 624 (42%)
Maximum Gain (bp): 63.6
Maximum Loss (bp): 45.4

Rule # 2: 5-year OIS and 10-year Gsec

Data: Daily closing yields from July 23, 2001 to April 29, 2009

Cross-Correlation Structure Suggests Positive Follow-through from Synthetic to Cash

Correlation (OIS (-1), Gsec (0)) = 0.14
Correlation (OIS (0), Gsec (0)) =0.48
Correlation (OIS (0), Gsec (-1)) = 0.03

Trading Rule

If Delta (5-year OIS Rate) > 0, then short 10-year government bonds
If Delta (5-year OIS Rate) <0, then buy 10-year government bonds
Delta: Daily close to close
Trade: Daily close to close


Performance Statistics*

Mean (bp): 0.88
Standard deviation (bp): 6.9
t-statistic: 5.4
No. of Positions: 1801
No. of loss-making positions: 751 (42%)
Maximum Gain (bp): 76.5
Maximum Loss (bp): 79.9

* Performance statistics not adjusted for the duration of the security traded.
Source: Bloomberg; IMF Working Paper: Derivatives Effect on Monetary Policy Transmission

SHORT-END OF THE CURVES COMPLETELY OUT OF LINE WITH THE CYCLICAL REALITY; 2-YEAR GSECS OFFER PLUS 20% ROE POTENTIAL WITH 10:1 LEVERAGE OVER 1-YEAR

The short-end of both the OIS and the risk free curve is completely out of line with the cyclical reality. While this discrepancy has corrected quite a bit since April 5, 2009 (when I first highlighted the tremendous return potential), the market still seems to be discounting a pretty sharp reversal of monetary policy.

1-year OIS is trading at 3.77%
1-year OIS 1-year forward is trading at 4.62%
1-year Gsecs are trading at 4.19%
1-year Gsec 1-year forward is trading at 5.16%

While it is difficult to make a case for further rate cuts from current levels, I am prepared to bet that if there is a move over the next-year, it will be lower. And I remain confident that the RBI will keep the market flushed with liquidity. The call, therefore, will hug the reverse repo rate (on average) over the next one-year.

For the year starting May 2010, the outlook is slightly murky. Yet, at this stage I doubt that short rates will climb 150bp higher from current levels in a relatively short period of time. The market is pricing that the RBI will normalize policy quickly once signs of a rebound are well entrenched. I doubt that. Policy makers globally are unlikely to risk nipping an incipient recovery in the bud.

Given the above, I would buy 2-year Gsecs with 10% equity for 1-year. I expect this trade to earn a return on equity of more than 20%.

GREEN SHOOTS AND CHERRY BLOSSOMS

NOT A GREAT DEAL OF DIRECTIONAL CONVICTION FROM HERE; CHASE THE RALLY VIA U.S. CREDIT

I missed the fast and furious global equities rally that started six-week back. While I do not have a great deal of directional conviction, I think risk-reward for jumping into the market at current levels is not great. Continued positive economic surprises would be harder to come by as economists revise forecasts based on the better than expected data flow. Even if good news graces the market, it is no longer oversold to cause the price action evidenced in March. On the other hand, investors could severely punish negative developments.

Yet, I continue to be long spread products: high-yield (HYG), high-yield loans (FFRHX) and investment-grade bonds (LQD). Credit has lagged equities and if this economic improvement is for real, there would be a sharp catch up rally. Battered valuations would limit downside in the event of a setback.

BANKING SYSTEM ISSUES: POTENTIAL FOR UPSIDE SURPRISES LIMITED GOING FORWARD: LONG = GOLDMAN SACHS AND MORGAN STANLEY; SHORT = CITI AND BANK OF AMERICA

The surge in equities began with the assertion of return to profitability by Citi and Bank of America. The earnings out so far have been better than expected, but the massive rally in financials (XLF: 77% up since March 9, 2009) has probably already discounted this outcome. Concerns remain (Goldman’s missing month, AIG payouts, concentration of revenues in trading, impact of the accounting rule change, credit issues in the loan book etc) and sustainability is the key question.

Investors continue to evaluate the likely outcome of the tug of war between a better competitive landscape and legacy assets. Lehman and Bear Stearns’ demise has left future profit streams (and talent to capitalize on this opportunity) on the table. In my view, strong securities focused franchises are in a better position to attract this talent and gain market share. The banks with huge loan books will continue face problems and this should become evident at some stage during the rest of the year (unemployment lags the cycle).

The next big event is the result of the stress tests. Debate rages on how the government should publish the outcome. Questions are already being raised about the reasonableness of the economic assumptions underlying this examination of the financial system. Any declaration of universal health will strip the exercise of credibility. The message so far is that the government will probably release bank level data with a plan for capital injection in some form for the institutions deemed weak. Whatever the outcome, it seems to me that disappointment is more likely than not. Also, the potential for negative surprises is higher for commercial banks with huge loan books.

The last factor is the success of the PPIP. My initial conclusion is that the program will be more potent in tackling the legacy securities issues. Loans are carried at amortized cost and even 6:1 leverage might lead to prices which might result in significant write-downs.

BEST HINDSIGHT TRADES OF Q1 2009: LONG EMERGING MARKETS, LONG USD/YEN, SHORT U.S. FINANCIALS AND SHORT LONG-DURATION U.S. TREASURIES

The following are the key trends from a survey of global asset class performance year-to-date:

1. Emerging market bonds as well as equities outperformed. The BRICs markets are back.India did well despite the general election uncertainty.
2. Duration exposure in the U.S. treasury market did not reward given significant supply issues, but the TIPs market surged as break-evens normalized.
3. The commodities downdraft seems to be over, suggesting that the global economy has bottomed and the inventory adjustment process is near its end.
4. The U.S. dollar gained ground in 2009, building further on the rally that started in 2008. The yen performed poorly as Japanese export concerns finally caught up with the currency.
5. Gold remains roughly unchanged buoyed by the twin forces of risk-aversion and the likely inflationary impact of the global policy moves.
6. In line with the overall emerging market resurgence and commodity revival theme, materials led sectoral performance in the U.S. Not surprisingly, financials lagged.

Global Asset Class Performance (2009 YTD)

Source: Market Watch

RISK-REWARD GREAT FOR DURATION EXPOSURE THROUGH HIGH QUALITY CORPORATE PAPER

I have been part wrong and part right as far Indian fixed income markets are concerned. 

DURATION EXPOSURE DID NOT PAY OFF IN Q1 2009; THE SHORT-END LOOKS ATTRACTIVE

I underestimated the brute force of supply pushing government security yields higher. Overnight indexed swap rates climbed as market-players probably paid to hedge duration. Yields rose on corporate paper as well.

Yet, the current yield levels – particularly the short-end - seem totally out of line with the cyclical reality, even after accounting for significant fiscal stimulus led issuance. Persistent excess liquidity is likely to be the norm going forward. I am also confident that the Reserve Bank is unlikely to lift the reverse repo rate over the next year. Given this backdrop, the short-end seems very attractively priced.

Consider the following investment-buy 1-year Government securities financed in repo with a haircut of 10%. Based on the current 1-year spot rate (5.35%) and daily call funding at 3.5% over the next 1-year, the return on equity on this investment will be 22%.

On the OIS front, the 1-year spot rate 1-year forward is 4.78% currently. This implies almost a 100bp increase in the 1-year spot OIS over the next 12-months – a scenario that seems extremely unlikely to me. A clear trade here is to receive the 2-year OIS versus the 1-year (receiving the 1-year OIS, 1-year forward).

SPREAD DURATION EXPOSURE TO HIGH QUALITY CORPOARTE PAPER REWARDED HANDSOMELY; ABSOLUTE YIELD LEVELS VERY ATTRACTIVE

The yield on the FIMMDA Bloomberg 10-year AAA corporate bond index rose about 55bp point in the first quarter of 2009, but the spread over 10-year Government security yield declined to 172bp on April 3, 2009 from 284bp on December 31, 2008. This spread averaged 111bp since December 2001, with a standard deviation of 76bp. At this point, the risk-reward for spread duration exposure does not look that great, but absolute yield levels on a diversified basket of corporate paper look attractive. Clearly, this is the time to assume duration exposure through high quality corporate paper.

RBI ANNOUNCES PLAN TO BUY INR 800BN GOVERNMENT SECURITIES

The overall structure of interest rates cannot decline unless base risk free yields fall. And, RBI’s latest move indicates that the central bank is alive to this issue. From the perch of current yield levels, long duration exposure makes sense. The ultimate aim of policy-makers is to reduce borrowing costs for the industry and that is where long positions would be most rewarding. 

SHORTING GAMMA IN THE USD/INR OPTIONS MARKET

The premise for short gamma trades is that the options’ market consistently overprices the volatility of the underlying. This discrepancy can be interpreted as a risk premium that speculators demand from hedgers. 

A simple test suggests that this hypothesis is true for the USD/INR options market. Also, this risk premium seems to be auto-correlated. 

RESULTS FROM IMPLIED VERSUS REALIZED VOL TRADES

I took one-week implied vols and compared these with USD/INR’s realized volatility over the subsequent seven-days. I calculated the daily standard deviation of the exchange rate on a close to close basis (annualized by multiplying it with the square root of the number of trading days per year). Then I measured the difference between implied and realized vol over the period under consideration. Following are the results from this exercise for non-overlapping weekly periods (annualized):

Data: 1-week implied vols. from July 2003 to March 2009; USD/INR: Daily data from July 2003 to March 2009 

 Mean: 1.6%

Stdev: 3.5%

No. of trades: 294

Proportion of loss-making trades: 22%

Maximum: 18.7%

Minimum: -10.3%

THERE IS EVIDENCE OF MOMENTUM IN WEEKLY PERFORMANCE (VOL TERMS)

The serial correlation is about 0.52, suggesting the presence of momentum. Based on this result, I tested a historical scheme to go short gamma only if the last trade money. The results are the following (annualized):

Mean: 2.0%

Stdev: 3.3%

No. of trades: 228

Proportion of loss-making trades: 17%

Maximum: 18.7%

Minimum: -6.2%

Clearly, the performance improves on many fronts. The proportion of loss-making trades decline by 5% to just 17% and the maximum drawdown falls as well.

 SLIPPAGES AND TRANSACTION COSTS HAVE NOT BEEN ACCOUNTED FOR AND COULD BE A PROBLEM

These numbers do not account for transaction costs and implicit to this scheme is delta hedging on a daily closing basis. There are bound to be slippages on this front. 

Bloomberg is the source of all data. 

 

REVIVAL IN GLOBAL RISK APPETITE IS KEY TO A SUSTAINABLE EQUITY RALLY AT HOME; RELATIVE INSULATION FROM GLOBAL WOES HAS NOT HELPED AND WILL NOT HELP

I have made the case in the past that Indian equities are cheap by historical norms, but still not at valuation extremes of December 2002. While current pricing is consistent with healthy returns over the medium-term (5-10 years), we are still not in an environment that would be conducive to a sustained revival in equity market fortunes. Indian indexes cannot and will not rally at variance with global risk appetite. Consider the following regressions:

Sensex Total Return (%) = 19% + 0.01 * U.S. HY Excess Returns

Data: 1990 to 2007; Yearly
U.S. High Yield Excess Returns: Return over duration matched treasuries in basis points on the Lehman Brothers U.S. High Yield Index

R-squared = 0.44
t-statistic (beta) = 3.58 (P-value:0.002)

Sensex Total Return (%) = 11% + 0.01 * U.S.$ EM Excess Returns

Data: 1993 to 2007; Yearly
U.S.$ Emerging Market Excess Returns: Return over duration matched treasuries in basis points on the Lehman Brothers U.S.$ EM Index

R-squared = 0.23
t-statistic (beta) = 1.99 (P-value: 0.06)

The key message is a positive relationship between Indian stock market variation and excess returns in U.S. spread sectors. When risk premium demanded by global investors climbs, all asset classes under-perform.

Another way to look at this phenomenon is that the Indian bourses will post significant and sustainable rallies only when foreign portfolio investors return in a big way.

I do not foresee a revival in global risk appetite in the near future. The incessant flow of bad news continues. Trouble in Eastern Europe, never ending issues in the banking system, and massive fiscal deficits; all continue to keep markets on the back foot.

All of the above does not rule out a 20%-30% rally from wherever the market bottoms in the current downdraft, but the sustainability of the upswing will be questionable and I will look to reduce risk as markets move higher.