I work in rates strategy for a top BB focusing on derivatives. Ask me anything on this topic including strategies, career advice, exit ops, education, derivative concepts, etc. I will give you the best answer to my abilities or direct you in a possible path to answer your question.
Products covered: swaps, swap spreads, swaptions, exotics. Broader team covers Treasuries, money market products, credit products, mortgage products, etc.
Clients: banks, HFs, AMs, pensions, insurance, sovereign, etc.
Main responsibilities: finding trading ideas for the clients, covering in depth analysis on fundamental drivers of the products mentioned above. Providing relevant commentary on moves in those markets as well as developing forecasts. Understand macro catalysts and their impact on the rates markets. Answering client/sales questions on relevant products, trading ideas, etc.
Division overview: sit on research desk but work closely with sales and trading. We develop the strategies for our clients, sales pitches our research/trading capabilities to our clients, and our traders execute and manage the risk whenever clients trade.
Background: I did a similar Q&A post when I worked as a quant analyst for 2 years here if you are interested in the quant space. My quant role involved derivatives pricing on the sell side after completing a masters in quant finance. Undergrad in pure math from a non-target.
Comps? What made you choose your current career pathway?
Standard comp at sell side. Anyone with institutional clients (sales, trading, research, and IB) typically all have the same base comp at the lower levels (but it could be different at other banks). After the recent inflation run up, they increased salaries so that Analyst 1 makes around 100k and then it typically increases 25k each year thereafter so Associates can make 150-200k ish. The bonus is where things will vary (for instance banking and trading typically have the highest although from this year alone banking bonuses in some groups I have heard were 0 and you'd be lucky for that as certain groups have been downsizing and outright laying people off). VPs can make >=225k and that is when the stock comp starts to kick in I believe. That is why being in research is the most stable in terms of compensation which may be less rewarding but you don't have to worry about managing P&L and the teams are lean so job security is good.
My pathway is slightly unconventional but basically my motto is do whatever it takes to climb up the ladder and get where you need to be but keep an open mind and be persistant. Then when you acquire an attractive enough skillset you can tailor it to get to exactly where you want to be. For instance, I started as a math major with little info about finance. I stumbled upon quant finance which got me in through the back door of finance with a masters degree. I was able to get a job as a quant (in derivative pricing you work with a lot of PhDs who are very smart but can continue doing that work for their whole career) and do the CFA. I had ambitions of transitioning and getting closer to markets which is what I am passionate about and I kept applying to the trading/strategy roles and landed the right opportunity - I received other offers but didn't settle because after 2 years of experience you kind of know what you want. Now I am in a position with a coveted skillset and can probably jump to the buy side when I feel ready (at the moment I don't want to because there is still plenty to learn in my current role).
so you have a phd?
Do you think rates follow a random walk? Or at least, a random walk with drift? If rates don't generate cash flow/ real yield, then what's the basis of mispricing in rates markets if not in relation to intrinsic value?
Coming from quant, what got you interested in rates? Is rates necessarily easier - quant is notoriously intense for obvious reasons
Oh and also, can a fundamental analyst succeed in rates? Or is it nearly all quant-ish now
Do rates follow a random walk? How do rate pricing models work?
For context and for those who are unfamiliar - random walk just says your next value depends on your previous value and drift means that the model should converge to some long term mean.
The answer depends on who you ask (someone in academia, a practitioner on the pricing side and a practitioner from strategy) and what their purpose is. Derivative pricing quants and researchers in academia will use models like random walks - they are called short rate models because they attempt to capture changes in rates for an infinitesimally short time period given some starting point. Some examples include Hull White, CIR, Vasicek among others. They believe rates are driven by some known process (drift) and some stochastic (random) process. The random process is usually driven by Brownian motion, which is a popular assumptions for equity (using Geometric Brownian motion). Sometimes these models will be run thousands of times in a Monte Carlo simulation and some average path might be used.
The problem is it’s these models - where do you start your curve? The answer is you calibrate your spot curve to the market and then derive forward rates from your model. Thus, your assumption is that the market is perfectly priced which we know is not true as interest rates move daily and market expectations change rapidly. It’s not a particularly good model for periods longer than a few days so that’s why these models are used for pricing rather than forecasting.
How do practitioners approach rate forecasts?
Forecasting is a totally different ballgame. There are no fancy models outside of regression (although PCA is common too). In this case you assume the current spot and forward curves are mispriced and you take advantage via relative value. It’s important to note that it is harder to make a prediction on outright yield levels (e.g. predicting where 10Y rates will be) — for this you need macro view on fundamental factors such as policy rates, inflation/growth expectations, liquidity premiums, evolution of the Fed's balance sheet, etc. (drivers will depend on which part of the yield curve you are forecasting) and empirical moves in these rates around similar macro regimes. It is much easier to take a view on the actual yield curve itself instead of predicting outright rates (e.g. predicting 2s10s versus predicting just 2s or just 10s). And it is even easier to take a view on rate butterflies versus rate curves (e.g. taking a view on 2s5s10s fly versus 2s5s curve or 5s10s curve). Typically you’ll have a fair value model say for the yield curve with a specific set of drivers and each will exhibit some beta with the curve, and if you have a view on those drivers going forward you have can form a view on the impact on the curve. On top of fair value, there are other typical regressions you can run to find some residual mispricing to typical relationships that you would normally expect to hold (i.e. you can run a curve vs rate regression and might expect yield curve to steepen in a rally and flatten in a sell off and if that is mispriced you can take advantage). Or for butterflies you can regress the fly against the belly rate and curve on the wings and take advantage of mean reversion in the fly relative to this relationship.
Cash flows of bonds/swaps
When you say rates don’t generate cash flow that is a false statement - of course they do! Just like other real assets there is an associated carry (coupon income minus financing cost) and roll down (P&L assuming an unchanged yield curve over some horizon). Bonds earn coupon income and you finance them in repo. Similar for swaps except your coupon income can be thought of whichever leg you receive in the swap (and there are no financing costs in swaps, only margining requirements).
What got me interested in rates?
It was the best opportunity at the time and given my derivative pricing background was rates related (which I happen to cover rates models in this role), this opportunity fell in my lap. And after a few years covering rates markets I love the fixed income markets given how closely they are related to macro and have mathematical complexity to them. Wouldn’t necessarily say it’s easier or harder than quant because it depends on your skill sets. Forecasting uses more statistics and time series analysis, whereas pricing uses more stochastic calculus. In a pure sense, "quants/desk strats" are typically for pricing but in a loose sense you need to be a "quant" for both areas as the term is can be very broad and just means someone who commands a good understanding of math/stats/finance/programming which are required for both roles.
Fundamental / equity analyst to fixed income
You can do anything you set your mind to so I'll start with that. However, I will say the thought process is totally different. Drivers in equity are driven by company fundamentals: it is a more bottom up analysis whereas rates markets are more top down on the macro front as mentioned earlier. Equity analysts typically only need just 10Y rate forecasts for their DCF and some basic macro assumptions to support their thesis for equity trades. But a constant discount rate is obviously an unrealistic assumption - we see that very clearly now in the most rapid tightening cycle in history. Historically with rates at zero lower bound it has been easy to ignore the rate factor in your model but everyone now is thinking of rates given the regime change we experienced and people's models breaking. Typically rates people are also a lot more quantitative vs. qualitative as well and getting used to fixed income products can be a challenge for people who haven't studied them or have practical experience in them. So if you read a lot of primers/textbooks on fixed income it could help but a lot of books are too academic and don't take into account a practitioners point of view so reading them won't really make you a better rates trader. Some practical books I recommend are the Treasury Bond Basis by Burghardt and Belton and Interest Rate Markets by Siddhartha Jha.
Outstanding and meticulous answer! Thank you!
Let me read this more carefully in my down time and circle back if I have any questions. Thanks!
For someone like me, early in their career, who got placed on a muni desk in an asset management/buyside type of role at a BB but has always been set on going into macro (with hopes of macro PM one day), what would your suggestions be on pivoting from such a role to macro strategy? I understand this may not be enough detail to adequately address the question so would love to PM if you were open so as not to dox myself.
Do you have thoughts on the MBA vs MS/MA Econ debate on breaking into macro strategy?
How would you recommend novices approach macro idea generation?
Do you have any recommended books or websites that you have found helpful to developing a better macro understanding?
How do you feel about trading vs strategy in preparation to move to a macro hedge fund? Also, what would be your advice to someone starting in a rates/fx sales position full time who eventually has the long term goal of macro pm? Finally, within macro S&T/research, how do you see the potential for career progression within a bank, and automation affecting each job? Is sales/strategy/trading more or less likely to be affected by automation? What do exits outside of macro HFS look like for sales, trading, and research roles in macro S&T? Finally, is there any potential for a non stem major to work in your role, or would I need more of a quantitative background? I have light python/R experience but am from an Econ BA. Thanks!
Research to buy side
Wondering this myself as well as I am still in the early stages of my rates career (only a few years in relative to MDs and PMs that have been in it for 20+ years who are pure experts). If someone else can chime in that would be helpful. On the rates strategy side, I am servicing clients and HFs so a lot of the trades we put on are for the clients so that would certainly be helpful in being able to take a view on products that hedge funds trade. I suspect a good exit op to a macro hedge fund would be to transition into an analyst for the PM and provide the trade ideas similar to what I do for the MD now. The downside in research is that you don't have a real book of P&L (we monitor our P&Ls and take our idea generation / trade construction very seriously) but it is not as rigorous as actual traders who literally have to manage the dollar risk & Greeks (for derivative books) daily is something beyond what you can learn in research/strategy. A good PM on the buy side for a macro fund will likely be able to generate ideas and manage the risk I imagine.
Sales
Sales is a much different role. The rates salespeople are very smart - the read the our research carefully, understand why we have our views and explain them on a basic level to the clients, pitch the business (research/trading capabilities), execute trades, and do a lot of other operational tasks / ad hoc client requests. Perhaps their best qualities lie in their ability to talk and work well with the clients. However, oftentimes they struggle to understand the finer details of our views or our trades or technical concepts on rates and they'll very frequently ping research with questions or defer technical client questions to us. Like I said earlier, anything is possible you set your mind to as long as you put yourself in a good position, network with the right people, be persistent and have a bit of luck. Although for macro PM on the buy side I assume the typical route would be sell side trader --> macro fund analyst --> PM or in my case research --> sell side trader which then will lead to buy side. I know people who have gotten to the trading desk from research so it is not uncommon, just need some time to pick up the knowledge and prove you're ready to handle the trading side. Coupled with research + trading knowledge will best equip you for a role on the buy side. Other exit ops are probably starting your own fund or transitioning to a different asset class I guess. Let me know what others think about other exit ops besides buy side analyst/trader/PM.
Automation
Given sales is the most operational of the three, I can see sales being automated first although managing relationships with people is something that is likely not going away anytime soon. In fact, in that case maybe trading in terms of the actual execution although traders have a lot of discretion even on the sell side to manage the risk or take delta leans. So likely there might be parts to trading that get automated but traders will still be needed so maybe just teams will downsize but it would be impossible (anytime soon) for trading to be automated. Research is possible the least affected by automation: research is all about idea generation, being creative, and being forward looking. AI is trained on historical data so if research was recycled from previous times in history it wouldn't be useful. In conclusion, roles in the investment bank are the least affected versus middle and back office roles, those will definitely be automated before sales/trading/research get automated.
Education
Best and fastest way to get into rates on the markets side is to either be an Ivy league undergrad (some traders get jobs right out of Harvard) or do what I did and get the quant finance masters and eventually get on a rates desk. Otherwise people with too much experience likely have missed their opportunity if they are in a different asset class or have graduated many years ago. Learning Python is beneficial for any role for an investment bank but might not be useful for some roles. Anything rates related it will be a must. If you are in pricing you'll need more complex languages like C++. I suggest you pick up quant topics (math, stats, programming, and financial engineering concepts) on your own whether it be from outside sources or further formal education if you want to break in and don't meet the other criteria.
Any idea of getting into rates research out of undergrad. Currently an applied math + econ student at a top school. How hard is it/what does it take?
How quant heavy is your area? Could I make it in with an UG degree in finance and not much math knowledge past Calc 2?
Hey read the above answer in the Education section I posted. In addition to that, I would say you'll need more than calc 2 for sure on rates but not random upper level math like proofs which are less useful. Mostly the upper levels that are applied math areas: stochastic calculus, time series analysis, numerical methods, modern portfolio theory, partial differential equations, linear algebra, probability theory, stats, and basic calculus (1-3). The first 4 in that list I learned in my masters and the others you typically learn in undergrad. Even now as practitioners we use a lot of random concepts from these fields. For instance PCA models use a lot of linear algebra or regression/time series models that use a lot of statistics and you need to be able to interpret these concepts intuitively and be able to explain them to clients if you use them in your research.
Labore rerum id aut. Est iure esse fugiat accusamus. Autem aliquam voluptatem veniam totam voluptatem nobis. Libero nisi nam nobis quia.
Quis nihil blanditiis dolor voluptatum non enim et. Itaque omnis nam labore aliquam nam aliquam aut. Ut facilis doloremque est at culpa.
Voluptates sint non architecto nulla qui dolores alias. Harum ut ipsam qui. Accusamus consequuntur et et consequuntur occaecati fuga. Eum laboriosam commodi aut error eligendi et. Quis molestiae consequuntur ut et.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...