macro paper trading

Breaking down each bond into coupons and principal in order to discount each flow at the correct rate like the textbooks say just to be theoretically correct and account for stuff like higher coupon bonds having lower yields under steep curves seems rather futile when in reality high coupon BTPs actually had higher yields and sometimes results of the fitting experiment aren’t even repeatable (due to different starting point). Working with YTMs directly is a lot more convenient. I tried to build the curve for various points between 2009 and 2014, and for last Friday. The results are entirely different from Bank of Italy and Bloomberg curves. The good thing is that there no sharp jumps in coefficients and no stuck points, at least if I stick to the same starting point in each experiment.

For a few semi-random time points picked the curves were quite smooth and residuals were on average 3-8 basis points except 2011H2. Early December 2011 was particularly ugly with bumps in the curve, 18 basis points average residual and entirely different very long-end model yield based on the fitting process starting point.

For now the BTP futures CTD yield is 8 bps above the model value.

uc


uc
 
10/02/2017 – 01/03/2017, short Fed Funds Apr17, +8882 USD
10/02/2017 – 10/03/2017, short EURJPY, +1156 USD
10/02/2017 – 08/03/2017, long BTP against Bund (March), +4370 EUR
10/02/2017 – 08/03/2017, short Schatz against ER Z7m -8773 EUR
10/02/2017 – 10/03/2017, short AUDCAD, -2030 USD
10/02/2017 – 31/03/2017, EDU1-EDU2, -2250 USD
10/02/2017 – 31/03/2017, long IR U8, +3337 AUD
15/02/2017 – 31/03/2017, long Short Sterling Sep 17, +240 GBP
21/02/2017 – 31/03/2017, steeper CAD, flatter AUD, -4388 CAD, +1554 AUD
09/03/2017 – 31/03/2017, long BTP against Bund (June), +970 EUR
10/03/2017 – 31/03/2017, short USDRUB, -588 000 RUB
13/03/2017 – 31/03/2017, ER Z7-Z8-Z9 fly, -2250 EUR
20/03/2017 – 31/03/2017, long BA Z8, +2250 CAD
20/03/2017 – 31/03/2017, long Bund against Long gilt, +3740 EUR, -3020 GBP
22/03/2017 – 31/03/2017, 2250-2400 long strangle on S&P, -1475 USD
28/03/2017 – 31/03/2017, short EURNOK, +2595 USD
By currency: +6872 USD, -1943 EUR, +4890 AUD, -2137 CAD, -588 000 RUB; total: -13 000 in USD terms.

16 trades, 7 losing ones. I will have to think about this.
 
It’s a curious case of wheat import tariffs in Turkey. On 15/03/2017 there was a rumour that Turkey slapped a 130% tariff on wheat imports from Russia that was subsequently denied on 17/03/2017

http://uk.reuters.com/article/turkey-wheat-russia-idUKL5N1GU3TY

There was another article today of Russia promising retaliatory sanctions, and I realized missing an article from 20/03/2017 that denied the denials of the tariff.

TRY CPI was released today with more than 1 percentage point increase to 11.3% yoy, but this was definitely expected given 1) no move in USDTRY, 2) central bank did say in March we will see more inflation in the near-term (and hiked too). The central bank was expecting that upmove on lagged FX effect and higher yoy food prices (base effect). I don’t have the recent weights for the CPI, but older weights show that food is 22% of the CPI basket with just wheat and bread accounting for 2%. Turkey imports 30% of wheat needs according to some article I read, and almost all of it comes from Russia. Losing 30% of wheat supplies with food accounting for a fifth of the CPI basket should manifest somehow in headline inflation. Regressions of CPI or even food CPI on wheat prices showed no strong R^2, so this is theory without much empirical backing (possibly food controls, I don't know). If we assume a 25% increase in wheat prices with a quarter of it passing to other food prices, we’re getting a two percentage impact on headline CPI and half percentage points with just wheat and bread being impacted (2% of the basket) and no pass-through to other food components.

April is seasonally inflationary month, so with an average April jump the yoy CPI will slightly accelerate from 11.3% to 12%. If we add the impact from food prices, it could look ugly. Food CPI itself is likely to accelerate anyway without any mom changes in food from 12.5% to 14.20% in April. Econ data hasn't been too bad, so no reasons to assume inflation subsiding a la Brazil on weaker demand.

USDTRY and EURTRY have been rather stable since the news story initially appeared on 15/03/2017, so this doesn’t seem to be priced in. I like to compare TRY with ZAR and BRL but wouldn’t want to look at those two due to Zuma and beef. This year it’s appreciated by (very roughly) 5% but it’s been mostly EM strength. On the other hand, TRY has been more than 20% down against both USD and EUR since mid 2016H2, around 30% against ZAR, so it’s a fairly beaten currency. I cannot recall if central bank has commented on its previous forecasts from the January inflation report, but they’re way off the 8% target for 2017. Not sure they will be within the upper band of the forecast of 9.4%. I am getting 9.64% by assuming average monthly gains for April-December. With wheat prices rising it could get a lot worse.

uc
 
Weekly:

EDU1-EDU2: +1500 USD. The markets did not react to balance sheet mentioning in the Fed minutes, so it’s too early to increase this yet.

Strangle on S&P 2250/2400: -362.50 USD

Short EURNOK: -832.5 USD

Long Short Sterling Sep17: +720 GBP. It’s time to close this position. Economic data hasn’t warranted rate cuts, and hiking odds are too unattractive to keep holding it.

Long bund against long gilt: +2860 EUR, -980 GBP. This weekly gain seems due to bunds richening against swaps and not due to EUR rates moving more than GBP rates. I see 10y swap differential is relatively unchanged at -40 bps, bund 10 bps down and 10y EUR swap rate 5 bps down. With the risk of getting screwed by benchmark change again, 10y bunds yield 50 bps lower than swaps, quite low historically. On the other hand, I have got burnt with rich schatz earlier in the topic, and bunds don’t look as expensive as schatz as those yield 63 bps less than swaps (throughout 2015 and 2016H1 the swap spreads were the same for schatzes and bunds).

Long BTP against bund: -600 EUR. Flat on the 1x position from 08/03/2017 (actually 4 bps narrowing but bund has higher duration apparently, so no gain), caught 15 bps narrowing on CTDs on the extra lot from 20/03/2017. They are 200 bps wide. Model shows 150 bps would be fair, but realistically will close the extra lot at no lower than 190 bps, all position somewhere between 170 and 185.

Euribor Z7-Z8-Z9 fly: +375 EUR. This one doesn’t want to perform for some reason. Draghi repeated no hikes before QE is over, so I want a lot lower Z7-Z8 relative to Z8-Z9. Both legs are 5-6 bps off peaks, and flatter curves are bad for me in this position (but good on portfolio level with long bunds and stirs in CAD, AUD).

Long BA Dec 18: +500 CAD

Steeper CAD/flatter AUD: -2463 CAD, +1940 AUD

Long IR Sep 18: +3288 AUD. Decided to pass taking profits at 98 for this position because Eurodollar curve shifted down too, so 90 day bills still seem too high relative to Eurodollars. On the other hand, hike prospects in the white strip are almost totally priced out, so this suggests quite limited downside for 90 day bill yields. Missed opportunity to close those at 98.06 on Friday prior to the afternoon sell-off.

Long USDRUB: +161 100 RUB. I will be closing this immediately on Monday if the opportunity is still there. Oil moved higher but RUB got punished by tomahawks, so RUB is no longer expensive versus oil.

Total: 305 USD, -260 GBP, +2635 EUR, -1963 CAD, +5228 AUD, +161 100 RUB; total of 6882 USD
 
short 250k USDCZK 3m NDF @ 25.081 spot (24.88 fwd)
short 2 6EM7 @ 1.06125
closed RUB short (180 SIM7 @ 58205)
closed 4 IRU8 @ 98.05 (still long 4)
 
sold 3 S&P 2250 puts @ 20.50 and 2 S&P 2400 calls @ 20.25. VIX has climbed more than 2 points from lows, so good opportunity to get rid of this position.
 
This treasuries rally this week is sad. I was contemplating adding to short long gilt through UST-gilt as my model shows it's got 10-30 bps too wide in the past few months, but the recent moves mean it's only fair to 20 bps now. FT article this week explained gilts being expensive with the usual boogeyman (brexit), the usual scapegoat (QE) and a fresh insight of pension funds putting their money to use by the end of the financial year, comments suggested it's lack of GBP paper supply as British companies prefer to issue in EUR.

Shorting 10 years outright is too early. In 10y USD swap slid about 35 bps down from the local peak prior to the Fed meeting, which historically is not an extreme. Sliding 20-25 bps more would make it a historically top quartile move and a better place to short outright. That would put us around 200 bps, still above pre election level, <redacted because everyone is sick of discussing Trump fiscal plans>.

Fed communications is a Fed trillemma to me. They 1) may change portfolio policy by year-end, 2) do not want to change portfolio policy before rates are normalized, 3) forecast fed rate increases for the next year. I'd expect the long end to be hit the most by fed unloading the portfolio. Out of 2.4t assets the Fed has: 2017-2019 maturity (38%), 2020-2022 (26%), 2023-2024 (7%), 2025-2027 (4%), 2028+ (24%). The treasury figures show that composition is currently: 2y (38%), 3-5 (29%), 5-7 (11%), 7-10 (8%), 10+ (13%). Fed is thus overweight 10y+, but underweight 7s and 10s. This is against the belief I previously had that Fed was heavily overweight 5+ so that balance sheet reduction would lead to 5s10 widening. The underweight 7s and 10s has naturally made me look at 2s10s30 and 5s10s30 but both look fairly directional, so have to refine that later.

Looking at USD swap rates from another perspective makes me doubt it really has 25 bps more to go. If we get 25 bps more in USD swaps, the pass-through to EUR paper would be likely to bring 10y bunds way below zero. I often look at 10y yield as a spread above 2y, 5y or whatever, but I think very many people look primarily at the absolute yield and feel uncomfortable about zero. I am somewhat sceptical of 10y bunds going way below zero because the stronger growth we're seeing in the eurozone should make deterioration in core inflation a lot less likely now, so incentives to hoard this for the long-run diminish. The governing council is likely to be polarized now. They primarily wanted to see wage inflation, and the data we had last month was more EZ & Germany labour cost growth and flat Italy after years of decline but slowdowns in France and Spain.
 
About flat for the week:

EDU1-EDU2: +1500 USD
Short EURNOK: +2288 USD
Long L U7: -250 GBP (closed)
Long bund against short long gilt: -1660 GBP, +1380 EUR
Long BTP against bund: -2960 EUR
ER Z7-Z8-Z9 fly: 0 EUR
Steeper CAD, flatter AUD: -4000 CAD, +1712 AUD
Long IR U8: +480 AUD
Long BA Z7: +1500 CAD
Short EURCZK: -130 EUR, -70 USD
Long S&P 2250/2400 strangle: +200 USD (closed)
Short USDRUB +220 USD (closed)
 
You must log in or register to reply here.
Back
Top