# Tag Info

28

Here are a few risks when using historical data: Data fidelity: Is your data an accurate reflection of history? For stocks, should you use actual closing prices or adjusted prices? For futures, how should you construct a realistic, continuous contract? Simulation realism: Are you making realistic assumptions about trade execution? Are you naively assuming, ...

16

This is practically a textbook case begging for the Kelly criterion. In your specific example, the optimal trade size is $f^*A$, where $f^*$ maximizes the average rate of return $$\mathbb{E}[\log (X)]=0.5\log(1+0.3f)+0.5\log(1-0.23f).$$ Here $f$ is the fraction of the current capital to trade. A straightforward calculation yields that ...

15

ML/AI systems are susceptible to a number of risks not traditionally discussed in risk management: What I call 'backtest arbitrage'. In the process of automated model generation and testing, your machine learner may discover, exploit, and concentrate on irregularities in your backtesting system which do not exist in the real world. If, for example, your ...

11

I assume you mean risk neutral pricing? Think of it this way (beware, oversimplification ahead ;-) You want to price a derivative on gold, a gold certificate. The product just pays the current price of an ounce in $. Now, how would you price it? Would you think about your risk preferences? No, you won't, you would just take the current gold price and ... 11 Accuracy The trader must make sure the data is not only right, but that the timestamps are useable. That's why a good data warehouse will be bitemporal or point-in-time. Thus, we know not only when the item was announced, but when we received it and could act on it. Gaps An aggressive safety check on incoming data might inadvertently exclude correct data. ... 9 Knowledge leads to profit BUT NOT the reverse The scientific method is your friend here. Wandering away from reality into fantasy is so tempting but will lead to failure. Step away from the P&L! Question -> observe -> theory -> predict -> measure -> record/publish/peer review -> repeat Think long and hard about what you believe are the facts/processes ... 9 Suppose that you and other bettors participate in a lottery with$N$possible outcomes; event will occur with probability$\pi_n$. There are$N$basic contracts available for purchase. Contract$n$costs$p_n$and entitles you to one dollar if outcome$n$occurs, zero otherwise. Now, imagine that you have a contingent claim that pays a complex payoff based ... 9 We bet on a fair coin toss -- heads you get$\$100$, tails you get $\$0$. So the expected value is$\$50$. But it is unlikely that you'll pay $\$50$to play this game because most people are risk averse. If you were risk neutral, then you WOULD pay$\$50$ for an expected value of $\$50$for an expected net payoff of$\$0$. A risk neutral player will accept ...

9

I'm not sure about the "CAPM formula" that you are referring to. I assume you are referring to the estimated coefficient of a regression of a security on a market portfolio. That is to say $$\beta_{security,market} = \frac{\sigma_{security,market}}{\sigma^2_{market}}$$ The idiosyncratic risk is the portion of risk unexplained ...

8

The risks involved in trading is everywhere and always a multifaceted thing: it includes the volatility of the selected asset, the leverage and concentration of the porfolio, whether there is a stop loss, a hedge, etc. Also, risk management is frequently not tied to the "alpha model" directly (e.g. VaR, shortfall, and scenario testing). For instance, one ...

8

I think the biggest risk is trusting your model too much. I would summarize modelling like that: Model for the best but risk-manage for the worst! As an example for modelling a portfolio approach with derivatives that could e.g. mean: use black scholes for option pricing (model) but manage your risk by assuming a power law distribution and vary your alpha ...

7

Risk Parity is not about "having the same volatility", it is about having each asset contributing in the same way to the portfolio overall volatility. The volatility of the portfolio is defined as: $$\sigma(w)=\sqrt{w' \Sigma w}$$ The risk contribution of asset $i$ is computed as follows: $$\sigma_i(w)= w_i \times \partial_{w_i} \sigma(w)$$ You can then ...

6

To give an example of a source of risk that isn't one of the ones you mentioned but still broadly on-topic for a Quant Finance site: operational risk - for which there are many references for contigency plans. This is the domain of the back office. Trades are created (priced and analysed) by quants, executed by traders and approved by preferably at least one ...

6

Each shop will differ - there is no widely used, unified framework shared across firms. Competitive advantages vary across shops, which ultimately reflect the biases/characteristics of the particular shop. Some will be far more mathematically sophisticated/inclined than others. Some maintain strong aversion to quantiative techniques such as risk models. ...

6

The shops I've worked for have had access to multiple brokers, but not for redundancy as your question implies. It's often because no one broker can handle every task. For example, I might need a floor broker, a dark-pool broker, an algo broker, and a separate prime broker. Each agency handles a different requirement. Even if one broker could handle all of ...

6

The Kelly criterion is a very popular bet-sizing method. Edward Thorp has written a great deal on this topic. You can try googling for more, or start with his review of the concept, or a recent paper, Medium Term Simulations of The Full Kelly and Fractional Kelly Investment Strategies. This is not specific to futures, but I'm not sure why you would need ...

6

The algorithm is certainly useful in that it is non-parametric, fast, and versatile. Meucci summarizes the advantages nicely: Unlike traditional copula techniques, CMA a) is not restricted to few parametric copulas such as elliptical or Archimedean; b) never requires the explicit computation of marginal cdf’s or quantile functions; c) does not ...

6

What you refer to as the 99.5th percentile is known as the "Value-at-Risk." You are correct that you will need to make a distributional assumption, and there is a popular and well-researched approach to this problem, though I'm not certain it could be called "standard." I would recommend you use the "truncated Levy flight" distribution. James Xiong at ...

6

I am a risk taker and I can say with confidence that you will never convince those individuals, you cited in your question, that they incur too much risk, because there will always be certain traders who prefer lottery tickets over longevity with the same firm (running high risk books unfortunately in the current environment runs equal to a free option; blow ...

5

Standard (read: regulators will accept it) could be a one day, 99% VaR calculated with two years of historical data. A minimum of one year of history is needed although this is not the norm. Typically the one-day VaR is transformed into a 10-day VaR by scaling the calculation by sqrt(10). However, the new market risk rule governs that one justify their ...

5

The risk-netural measure has a massively important property which is worth making very clear: The price of any trade is equal to the expectation of the trade’s winnings and losses under the risk-neutral measure. This property gives us a scheme for pricing derivatives: take a collection of prices of trades that exist in the market (eg swap rates, bond ...

5

There are all sorts of financial and non-financial risks. I define financial risk as all risks defined from events in the financial markets that affect all participants. Non-financial risks are all other forms of risk (including risks that a particular firm may face). Financial: Market value risk (interest rate risk, exchange prices, equity prices, ...

5

Since both $ER$ and $S$ are gaussian random, why not just assume their dependence is captured by their covariance, and make your draws from the bivariate normal distribution? It is hard to construct any other way of making two marginal gaussians cointegrated. Even if the variables were not gaussian, you would probably find yourself relating them using a ...

5

I am not sure why your question had so many upvotes because in currency markets anything else but triangular arbitrage does not exist. What is a quadrangular arb, I have never heard of it despite having traded fx among other asset classes for over ten years now. Think about it: Lets say you observe the price of EUR/USD. You can build triangular arbs by ...

5

Let's first restate the formula of the beta of a portfolio $P$ relative to a benchmark $B$: $$\beta_P=\frac{Cov(r_P,r_B)}{Var(r_B)}$$ As chrisaycock said in his comment, the key thing to understand is that the beta is a statistical measure computed relative to a benchmark. Hence, I believe that the real question you should be asking is: Which benchmark ...

5

Step 1: Get your data from SQL into R -> http://www.r-bloggers.com/?s=SQL Step 2: Run your analysis/optimizations like -> http://www.r-bloggers.com/portfolio-optimization-in-r-part-1/ or http://blog.streeteye.com/blog/2012/01/portfolio-optimization-and-efficient-frontiers-in-r/ or via RMetrics: ...

4

I consider market risk, credit risk and operational risk to be the three major forms of financial risk exposure. @jeebs addressed the trade settlement component of operational risk. I would also include the third bullet point that @shane gave in his answer as belonging to the category of operational risk. Another form of non-financial risk would be ...

4

Data is the lifeblood of a quantitative strategy. So I would say that the primary operational risks facing quantitative models are related to data. Some places where this can be an issue: Misinterpreting post-hoc data: Many economic indicators are revised on a periodic basis, and it's critical to understand what the meaning of the numbers are on a ...

4

I think, as with many machine learning approaches to investing decision support, it depends largely on the data. With a good selection of features, yes dynamic models like you're talking about will probably do better than a simple linear regression; but then again, with a good selection of features, linear regression will probably work reasonably well, too. ...

4

Typically these options are sold to portfolio managers to help smooth out their returns in times of trouble. A call (or even long position) on such an index will give a little PL in precisely the sort of markets that long-biased portfolio managers often lose money in, since high volatility is empirically correlated with negative returns. That keeps risk ...

Only top voted, non community-wiki answers of a minimum length are eligible