L6

Christoph Siebenbrunner:

Clearing Algorithms and Network Centrality

I show that the solution of a standard clearing model commonly used in contagion analyses for financial systems can be expressed as a specific form of a generalized Katz centrality measure under conditions that correspond to a system-wide shock. This result provides a formal explanation for earlier empirical results which showed that Katz-type centrality measures are closely related to contagiousness. It also allows assessing the assumptions that one is making when using such centrality measures as systemic risk indicators. I conclude that these assumptions should be considered too strong and that, from a theoretical perspective, clearing models should be given preference over centrality measures in systemic risk analyses.


Andreas Sojmark:

An SPDE Model for Systemic Risk with Default Contagion

In this talk, I will present a structural model for systemic risk, phrased as an interacting particle system for $N$ financial institutions, where each institution is removed upon default and this has a contagious effect on the rest of the system. Moreover, the financial instituions display herding behavior and they are exposed to correlated noise, which turns out to be an important driver of the contagion mechanism. Ultimately, the motivation is to provide a clearer connection between the insights from dynamic mean field models and the detailed study of contagion in the (mostly static) network-based literature. Mathematically, we prove a propagation of chaos type result for the large population limit, where the limiting object is characterized as the unique solution to a nonlinear SPDE on the positive half-line with Dirichlet boundary. This is based on joint work with Ben Hambly and I will also point out some interesting future directions, which are part of ongoing work with Sean Ledger.

 I will discuss my ongoing project towards a version of the Modular Andre-Oort Conjecture incorporating the derivatives of the j function.  The work originates with Jonathan Pila, who formulated the first "Modular Andre-Oort with Derivatives" conjecture.  The problem can be approached via o-minimality; I will discuss two categories of result.  The first is a weakened version of Jonathan's conjecture.  Under an algebraic independence conjecture (of my own, though it follows from standard conjectures), the result is equivalent to the statement that Jonathan's conjecture holds.  
The second result is conditional on the same algebraic independence conjecture - it specifies more precisely how the special points in varieties can occur in this context.  
If time permits, I will discuss my most recent work towards making the two results uniform in algebraic families.

In 2016, Habegger and Pila published a proof of the Zilber-Pink conjecture for curves in abelian varieties (all defined over $\mathbb{Q}^{\rm alg}$). Their article also contained a proof of the same conjecture for a product of modular curves that was conditional on a strong arithmetic hypothesis. Both proofs were extensions of the Pila-Zannier strategy based in o-minimality that has yielded many results in this area. In this talk, we will explain our generalisation of the strategy to the Zilber-Pink conjecture for any Shimura variety. This is joint work with J. Ren.

Komlós conjectured in 1981 that among all graphs with minimum degree at least $d$, the complete graph $K_{d+1}$ minimises the number of Hamiltonian subsets, where a subset of vertices is Hamiltonian if it contains a spanning cycle. We prove this conjecture when $d$ is sufficiently large.  In fact we prove a stronger result: for large $d$, any graph $G$ with average degree at least $d$ contains almost twice as many Hamiltonian subsets as $K_{d+1}$, unless $G$ is isomorphic to $K_{d+1}$ or a certain other graph which we specify. This is joint work with Hong Liu, Maryam Sharifzadeh and Katherine Staden.

A central theorem in combinatorics is Sperner’s Theorem, which determines the maximum size of a family in the Boolean lattice that does not contain a 2-chain. Erdos later extended this result and determined the largest family not containing a k-chain. Erdos and Katona and later Kleitman asked how many such chains must appear in families whose size is larger than the corresponding extremal result.

This question was resolved for 2-chains by Kleitman in 1966, who showed that amongst families of size M in the Boolean lattice, the number of 2-chains is minimized by a family whose sets are taken as close to the middle layer as possible. He also conjectured that the same conclusion should hold for all k, not just 2. The best result on this question is due to Das, Gan and Sudakov who showed roughly that Kleitman’s conjecture holds for families whose size is at most the size of the k+1 middle layers of the Boolean lattice. Our main result is that for every fixed k and epsilon, if n is sufficiently large then Kleitman’s conjecture holds for families of size at most (1-epsilon)2^n, thereby establishing Kleitman’s conjecture asymptotically (in a sense). Our proof is based on ideas of Kleitman and Das, Gan and Sudakov.

Joint work with Jozsef Balogh.

Many extremal results on cycles use what may be called BFS method, where a breath first search tree is used as a skeleton to build desired structures. A well-known example is the Bondy-Simonovits theorem that every n-vertex graph with more than 100kn^{1+1/k} edges contains an even cycle of length 2k. The standard BFS method, however, is not easily applicable for supersaturation problems where one wishes to show the existence of many copies of a given  subgraph. The method is also not easily applicable in the hypergraph setting.

In this talk, we focus on some variants of the standard BFS method. We use one of these in conjunction with some useful general reduction theorems that we develop to establish the supersaturation of loose (linear) even cycles in linear hypergraphs. This extends Simonovits' supersaturation theorem on even cycles in graphs. This is joint work with Liana Yepremyan.

If time allows, we will also discuss another variant (joint with Jie Ma) used in the study of Berge cycles of consecutive lengths in hypergraphs.

It is well known that there is a finite colouring of the natural numbers such that there is no infinite set X with X+X (the pairwise sums from X, allowing repetition) monochromatic. It is easy to extend this to the rationals. Hindman, Leader and Strauss showed that there is also such a colouring of the reals, and asked if there exists a space 'large enough' that for every finite colouring there does exist an infinite X with X+X monochromatic. We show that there is indeed such a space. Joint work with Imre Leader.

Kim's iterative non-abelian reciprocity laws carve out a sequence of subsets of the adelic points of a suitable algebraic variety, containing the global points. Like Ellenberg's obstructions to the existence of global points, they are based on nilpotent approximations to the variety. Systematically exploiting this idea gives a sequence starting with the Brauer-Manin obstruction, based on the theory of obstruction towers in algebraic topology. For Shimura varieties, nilpotent approximations are inadequate as the fundamental group is nearly perfect, but relative completions produce an interesting obstruction tower. For modular curves, these maps take values in Galois cohomology of modular forms, and give obstructions to an adelic elliptic curve with global Tate module underlying a global elliptic curve.

It is a standard heuristic that sums of oscillating number theoretic functions, like the M\"obius function or Dirichlet characters, should exhibit squareroot cancellation. It is often very difficult to prove anything as strong as that, and we generally expect that if we could prove squareroot cancellation it would be the best possible bound. I will discuss recent results showing that, in fact, certain averages of multiplicative functions exhibit a bit more than squareroot cancellation.